Effect on exit exercise decisions
outcome_cols = c('started_exercise' = 'Started exercise',
'completed_survey' = "Completed survey",
'redistribute_decision' = "Redistribute decision",
'dictator_decision' = "Dictator decision",
'lying_payout' = "Lying payout",
'trolley_decision' = "Trolley decision",
'skip_iat' = "Skip IAT score",
'iat_score' = "IAT Score",
'book_econ' = "Economics book",
'book_gender' = "Gender book",
'book_morality' = "Morality book",
'book_independence' = "Judicial independence book")
note <- "Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level."
# models_plot <- feols(fml = c(started_exercise, completed_survey, redistribute_decision,
# dictator_decision, lying_payout, trolley_decision,
# skip_iat,iat_score, book_econ, book_gender, book_morality,
# book_independence) ~ i(round, monitoreo) | idcurso,
# data = data,
# split = ~round, cluster = ~idcurso+dni)
for (outcome in names(outcome_cols)) {
formula <- as.formula(paste0(outcome, ' ~ monitoreo * teacher_female | idcurso | 0 | aula'))
out_list <- list()
for (r in 1:9) {
model <- felm(formula, data = data %>% filter(round == r))
out_list[[sprintf('Round %i', r)]] <- model
}
formula <- as.formula(paste0(outcome, ' ~ monitoreo * teacher_female | 0 | 0 | dni'))
out_list[['Pure Treatment']] <- felm(formula, data = pure_control)
out_list[['Perc. courses monitored']] <- felm(formula, data = monit_perc)
out_name <- outcome_cols[[outcome]]
cat('<h4></h4>')
cat(sprintf('<h3>%s</h3>', out_name))
stargazer(out_list,
covariate.labels = c('Monitoring', 'Female Teacher', 'Monitoring x Female Teacher'),
dep.var.labels = out_name,
column.labels = names(out_list),
notes = note,
omit = 'Constant',
type = 'html')
feols(fml = as.formula(sprintf('%s ~ i(round, monitoreo) | idcurso', outcome)),
data = data, cluster = ~aula) %>%
iplot(dict = outcome_cols, xlab = 'Round')
}
Started exercise
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Started exercise
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.033
|
-0.051
|
0.059
|
0.014
|
-0.009
|
-0.004
|
0.002
|
-0.004
|
-0.069
|
-0.026
|
-0.002
|
|
|
(0.042)
|
(0.067)
|
(0.040)
|
(0.056)
|
(0.051)
|
(0.053)
|
(0.054)
|
(0.053)
|
(0.079)
|
(0.041)
|
(0.037)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
-0.289***
|
-0.044
|
0.084**
|
0.033
|
-0.008
|
-0.058*
|
-0.048
|
0.135***
|
-0.100*
|
0.011
|
-0.006
|
|
|
(0.031)
|
(0.041)
|
(0.042)
|
(0.059)
|
(0.059)
|
(0.035)
|
(0.034)
|
(0.046)
|
(0.054)
|
(0.027)
|
(0.022)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.105
|
-0.160
|
-0.056
|
0.057
|
|
-0.009
|
|
0.121
|
-0.00000
|
-0.003
|
|
|
(0.000)
|
(0.099)
|
(0.150)
|
(0.106)
|
(0.117)
|
(0.000)
|
(0.124)
|
(0.000)
|
(0.088)
|
(0.038)
|
(0.034)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
558
|
558
|
558
|
557
|
536
|
534
|
534
|
529
|
532
|
3,550
|
4,845
|
|
R2
|
0.023
|
0.004
|
0.009
|
0.002
|
0.002
|
0.012
|
0.015
|
0.017
|
0.008
|
0.001
|
0.0001
|
|
Adjusted R2
|
0.018
|
-0.003
|
0.002
|
-0.006
|
-0.006
|
0.005
|
0.005
|
0.009
|
0.001
|
0.0003
|
-0.001
|
|
Residual Std. Error
|
0.376 (df = 554)
|
0.380 (df = 553)
|
0.379 (df = 553)
|
0.381 (df = 552)
|
0.385 (df = 531)
|
0.382 (df = 529)
|
0.382 (df = 528)
|
0.382 (df = 524)
|
0.383 (df = 527)
|
0.393 (df = 3546)
|
0.381 (df = 4841)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Completed survey
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Completed survey
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.012
|
-0.090
|
-0.003
|
-0.052
|
-0.072
|
-0.065
|
-0.043
|
-0.053
|
-0.115
|
-0.071
|
-0.063
|
|
|
(0.043)
|
(0.063)
|
(0.042)
|
(0.063)
|
(0.058)
|
(0.054)
|
(0.055)
|
(0.053)
|
(0.092)
|
(0.044)
|
(0.041)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
-0.355***
|
-0.095**
|
0.061
|
-0.013
|
-0.017
|
-0.082**
|
-0.102***
|
0.059
|
-0.120**
|
-0.008
|
-0.027
|
|
|
(0.029)
|
(0.042)
|
(0.049)
|
(0.075)
|
(0.066)
|
(0.038)
|
(0.036)
|
(0.044)
|
(0.053)
|
(0.030)
|
(0.025)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.154
|
-0.156
|
-0.004
|
0.095
|
|
-0.023
|
|
0.131
|
0.010
|
0.005
|
|
|
(0.000)
|
(0.108)
|
(0.180)
|
(0.102)
|
(0.116)
|
(0.000)
|
(0.132)
|
(0.000)
|
(0.095)
|
(0.042)
|
(0.039)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
558
|
558
|
558
|
557
|
536
|
534
|
534
|
529
|
532
|
3,550
|
4,845
|
|
R2
|
0.030
|
0.008
|
0.009
|
0.004
|
0.007
|
0.007
|
0.015
|
0.005
|
0.015
|
0.007
|
0.005
|
|
Adjusted R2
|
0.024
|
0.001
|
0.002
|
-0.003
|
-0.001
|
-0.0002
|
0.006
|
-0.003
|
0.008
|
0.006
|
0.004
|
|
Residual Std. Error
|
0.424 (df = 554)
|
0.429 (df = 553)
|
0.429 (df = 553)
|
0.431 (df = 552)
|
0.433 (df = 531)
|
0.431 (df = 529)
|
0.431 (df = 528)
|
0.432 (df = 524)
|
0.430 (df = 527)
|
0.428 (df = 3546)
|
0.429 (df = 4841)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Redistribute decision
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Redistribute decision
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.173
|
0.092
|
0.015
|
0.292
|
0.362
|
0.331
|
0.465**
|
0.319
|
0.194
|
0.419**
|
0.395**
|
|
|
(0.190)
|
(0.248)
|
(0.266)
|
(0.225)
|
(0.283)
|
(0.209)
|
(0.182)
|
(0.221)
|
(0.326)
|
(0.198)
|
(0.184)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.281
|
0.206
|
-0.253
|
0.362
|
0.319
|
-0.384**
|
0.905***
|
-0.156
|
-0.240
|
0.206
|
0.134
|
|
|
(0.272)
|
(0.375)
|
(0.335)
|
(0.365)
|
(0.363)
|
(0.171)
|
(0.164)
|
(0.127)
|
(0.221)
|
(0.139)
|
(0.122)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.091
|
0.561
|
-0.250
|
-0.175
|
|
-0.951***
|
|
0.152
|
-0.169
|
-0.108
|
|
|
(0.000)
|
(0.450)
|
(0.357)
|
(0.390)
|
(0.412)
|
(0.000)
|
(0.214)
|
(0.000)
|
(0.358)
|
(0.186)
|
(0.175)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
456
|
456
|
456
|
455
|
435
|
434
|
434
|
429
|
432
|
2,847
|
3,948
|
|
R2
|
0.004
|
0.007
|
0.008
|
0.008
|
0.011
|
0.022
|
0.029
|
0.018
|
0.005
|
0.013
|
0.010
|
|
Adjusted R2
|
-0.003
|
-0.002
|
-0.001
|
-0.001
|
0.001
|
0.013
|
0.018
|
0.008
|
-0.005
|
0.012
|
0.009
|
|
Residual Std. Error
|
1.777 (df = 452)
|
1.783 (df = 451)
|
1.782 (df = 451)
|
1.777 (df = 450)
|
1.783 (df = 430)
|
1.781 (df = 429)
|
1.769 (df = 428)
|
1.774 (df = 424)
|
1.793 (df = 427)
|
1.760 (df = 2843)
|
1.773 (df = 3944)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Dictator decision
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Dictator decision
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.119
|
0.307
|
-0.092
|
0.194
|
0.337**
|
0.187
|
0.069
|
0.062
|
0.118
|
0.101
|
0.106
|
|
|
(0.173)
|
(0.243)
|
(0.229)
|
(0.141)
|
(0.141)
|
(0.137)
|
(0.137)
|
(0.151)
|
(0.173)
|
(0.203)
|
(0.189)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.173
|
0.426
|
-0.309
|
0.346
|
0.509**
|
0.271**
|
0.706***
|
-0.037
|
-0.144
|
0.170
|
0.033
|
|
|
(0.118)
|
(0.419)
|
(0.209)
|
(0.358)
|
(0.234)
|
(0.124)
|
(0.096)
|
(0.139)
|
(0.215)
|
(0.152)
|
(0.131)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.601
|
0.568**
|
-0.299
|
-0.617*
|
|
-0.414***
|
|
-0.110
|
-0.090
|
0.046
|
|
|
(0.000)
|
(0.489)
|
(0.251)
|
(0.433)
|
(0.353)
|
(0.000)
|
(0.148)
|
(0.000)
|
(0.227)
|
(0.201)
|
(0.188)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
454
|
454
|
454
|
453
|
433
|
432
|
432
|
427
|
430
|
2,829
|
3,930
|
|
R2
|
0.004
|
0.008
|
0.008
|
0.006
|
0.010
|
0.004
|
0.009
|
0.0004
|
0.004
|
0.001
|
0.001
|
|
Adjusted R2
|
-0.003
|
-0.001
|
-0.001
|
-0.003
|
0.001
|
-0.005
|
-0.003
|
-0.009
|
-0.005
|
0.0004
|
0.0002
|
|
Residual Std. Error
|
1.839 (df = 450)
|
1.842 (df = 449)
|
1.843 (df = 449)
|
1.843 (df = 448)
|
1.876 (df = 428)
|
1.880 (df = 427)
|
1.877 (df = 426)
|
1.863 (df = 422)
|
1.884 (df = 425)
|
1.807 (df = 2825)
|
1.855 (df = 3926)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Lying payout
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Lying payout
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.0002
|
0.015
|
0.022
|
0.031**
|
-0.003
|
0.025
|
0.018
|
0.027
|
0.034
|
0.024
|
0.019
|
|
|
(0.025)
|
(0.037)
|
(0.022)
|
(0.016)
|
(0.033)
|
(0.028)
|
(0.029)
|
(0.029)
|
(0.022)
|
(0.032)
|
(0.028)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.138***
|
0.013
|
0.032
|
0.111**
|
-0.020
|
-0.010
|
-0.011
|
-0.038
|
0.117
|
0.029
|
0.021
|
|
|
(0.020)
|
(0.032)
|
(0.035)
|
(0.054)
|
(0.040)
|
(0.024)
|
(0.025)
|
(0.023)
|
(0.074)
|
(0.026)
|
(0.018)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.007
|
0.019
|
-0.069
|
0.052
|
|
0.045
|
|
-0.114
|
-0.003
|
0.003
|
|
|
(0.000)
|
(0.052)
|
(0.062)
|
(0.077)
|
(0.064)
|
(0.000)
|
(0.057)
|
(0.000)
|
(0.080)
|
(0.035)
|
(0.031)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
452
|
452
|
452
|
451
|
431
|
430
|
430
|
425
|
428
|
2,811
|
3,912
|
|
R2
|
0.009
|
0.002
|
0.005
|
0.016
|
0.004
|
0.003
|
0.004
|
0.004
|
0.013
|
0.002
|
0.002
|
|
Adjusted R2
|
0.002
|
-0.007
|
-0.004
|
0.007
|
-0.006
|
-0.006
|
-0.007
|
-0.005
|
0.004
|
0.001
|
0.001
|
|
Residual Std. Error
|
0.285 (df = 448)
|
0.286 (df = 447)
|
0.286 (df = 447)
|
0.285 (df = 446)
|
0.293 (df = 426)
|
0.293 (df = 425)
|
0.294 (df = 424)
|
0.295 (df = 420)
|
0.293 (df = 423)
|
0.338 (df = 2807)
|
0.291 (df = 3908)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Trolley decision
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Trolley decision
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.024
|
0.021
|
0.004
|
0.024
|
0.047
|
-0.015
|
0.005
|
-0.030
|
0.011
|
0.007
|
0.008
|
|
|
(0.038)
|
(0.042)
|
(0.040)
|
(0.048)
|
(0.056)
|
(0.039)
|
(0.043)
|
(0.037)
|
(0.044)
|
(0.049)
|
(0.045)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
-0.083*
|
-0.010
|
-0.016
|
-0.030
|
0.047
|
-0.059
|
0.026
|
-0.040*
|
-0.105
|
-0.008
|
-0.011
|
|
|
(0.049)
|
(0.071)
|
(0.065)
|
(0.066)
|
(0.060)
|
(0.040)
|
(0.040)
|
(0.021)
|
(0.091)
|
(0.032)
|
(0.026)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.006
|
0.053
|
-0.017
|
-0.087
|
|
-0.098
|
|
0.021
|
-0.016
|
-0.019
|
|
|
(0.000)
|
(0.082)
|
(0.094)
|
(0.069)
|
(0.062)
|
(0.000)
|
(0.060)
|
(0.000)
|
(0.096)
|
(0.044)
|
(0.040)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
451
|
451
|
451
|
450
|
430
|
429
|
429
|
424
|
427
|
2,802
|
3,903
|
|
R2
|
0.003
|
0.002
|
0.002
|
0.003
|
0.003
|
0.008
|
0.010
|
0.011
|
0.009
|
0.0003
|
0.001
|
|
Adjusted R2
|
-0.004
|
-0.007
|
-0.007
|
-0.006
|
-0.006
|
-0.002
|
-0.001
|
0.001
|
-0.001
|
-0.001
|
-0.0003
|
|
Residual Std. Error
|
0.420 (df = 447)
|
0.420 (df = 446)
|
0.420 (df = 446)
|
0.420 (df = 445)
|
0.423 (df = 425)
|
0.421 (df = 424)
|
0.421 (df = 423)
|
0.422 (df = 419)
|
0.421 (df = 422)
|
0.423 (df = 2798)
|
0.420 (df = 3899)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Skip IAT score
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Skip IAT score
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.087***
|
0.076
|
0.055*
|
0.060***
|
0.054**
|
0.092***
|
0.091***
|
0.079**
|
0.084**
|
0.117**
|
0.101**
|
|
|
(0.029)
|
(0.050)
|
(0.031)
|
(0.021)
|
(0.026)
|
(0.031)
|
(0.034)
|
(0.032)
|
(0.034)
|
(0.048)
|
(0.044)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.013
|
-0.004
|
-0.052
|
-0.060
|
-0.039
|
-0.059**
|
0.086***
|
0.142***
|
0.011
|
0.001
|
-0.003
|
|
|
(0.024)
|
(0.044)
|
(0.033)
|
(0.054)
|
(0.049)
|
(0.024)
|
(0.023)
|
(0.028)
|
(0.058)
|
(0.031)
|
(0.026)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.017
|
0.088**
|
0.122
|
0.123*
|
|
0.033
|
|
-0.013
|
0.035
|
0.039
|
|
|
(0.000)
|
(0.067)
|
(0.042)
|
(0.077)
|
(0.068)
|
(0.000)
|
(0.048)
|
(0.000)
|
(0.068)
|
(0.045)
|
(0.042)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
461
|
461
|
461
|
460
|
440
|
439
|
439
|
434
|
437
|
2,873
|
3,991
|
|
R2
|
0.012
|
0.011
|
0.015
|
0.017
|
0.020
|
0.017
|
0.030
|
0.023
|
0.011
|
0.022
|
0.015
|
|
Adjusted R2
|
0.006
|
0.002
|
0.007
|
0.009
|
0.011
|
0.008
|
0.018
|
0.014
|
0.001
|
0.021
|
0.014
|
|
Residual Std. Error
|
0.411 (df = 457)
|
0.412 (df = 456)
|
0.411 (df = 456)
|
0.411 (df = 455)
|
0.408 (df = 435)
|
0.409 (df = 434)
|
0.405 (df = 433)
|
0.408 (df = 429)
|
0.413 (df = 432)
|
0.420 (df = 2869)
|
0.408 (df = 3987)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
IAT Score
|
|
|
|
Dependent variable:
|
|
|
|
|
|
IAT Score
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.069**
|
0.101***
|
0.084*
|
0.099***
|
0.071*
|
0.078**
|
0.081**
|
0.097**
|
0.037
|
0.094**
|
0.089**
|
|
|
(0.033)
|
(0.036)
|
(0.046)
|
(0.035)
|
(0.041)
|
(0.032)
|
(0.032)
|
(0.038)
|
(0.039)
|
(0.044)
|
(0.041)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.006
|
0.009
|
0.020
|
-0.006
|
-0.062
|
-0.056**
|
-0.182***
|
-0.016
|
-0.083***
|
-0.042
|
-0.034
|
|
|
(0.047)
|
(0.053)
|
(0.052)
|
(0.047)
|
(0.050)
|
(0.025)
|
(0.017)
|
(0.034)
|
(0.027)
|
(0.031)
|
(0.025)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.073
|
-0.025
|
-0.126**
|
-0.003
|
|
0.178***
|
|
0.024
|
-0.005
|
-0.001
|
|
|
(0.000)
|
(0.077)
|
(0.068)
|
(0.060)
|
(0.062)
|
(0.000)
|
(0.035)
|
(0.000)
|
(0.045)
|
(0.040)
|
(0.037)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
292
|
292
|
292
|
291
|
277
|
278
|
278
|
274
|
275
|
1,754
|
2,530
|
|
R2
|
0.014
|
0.018
|
0.015
|
0.030
|
0.024
|
0.019
|
0.032
|
0.027
|
0.011
|
0.025
|
0.018
|
|
Adjusted R2
|
0.004
|
0.005
|
0.002
|
0.016
|
0.009
|
0.005
|
0.014
|
0.012
|
-0.003
|
0.024
|
0.017
|
|
Residual Std. Error
|
0.307 (df = 288)
|
0.306 (df = 287)
|
0.306 (df = 287)
|
0.305 (df = 286)
|
0.306 (df = 272)
|
0.307 (df = 273)
|
0.305 (df = 272)
|
0.307 (df = 269)
|
0.306 (df = 270)
|
0.301 (df = 1750)
|
0.304 (df = 2526)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Economics book
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Economics book
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.0002
|
0.005
|
-0.0003
|
0.005
|
0.043***
|
0.012
|
0.017
|
0.002
|
-0.036*
|
-0.008
|
-0.00004
|
|
|
(0.017)
|
(0.031)
|
(0.021)
|
(0.015)
|
(0.017)
|
(0.021)
|
(0.022)
|
(0.021)
|
(0.020)
|
(0.028)
|
(0.027)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.012
|
0.006
|
0.005
|
-0.017
|
0.028
|
0.082***
|
0.049***
|
-0.078***
|
-0.075***
|
-0.001
|
-0.002
|
|
|
(0.020)
|
(0.024)
|
(0.022)
|
(0.027)
|
(0.018)
|
(0.017)
|
(0.018)
|
(0.015)
|
(0.025)
|
(0.017)
|
(0.014)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.010
|
0.022
|
-0.012
|
-0.113***
|
|
-0.080*
|
|
0.058*
|
-0.002
|
-0.011
|
|
|
(0.000)
|
(0.048)
|
(0.026)
|
(0.050)
|
(0.022)
|
(0.000)
|
(0.045)
|
(0.000)
|
(0.034)
|
(0.021)
|
(0.021)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
439
|
439
|
439
|
438
|
419
|
418
|
418
|
413
|
416
|
2,721
|
3,801
|
|
R2
|
0.001
|
0.002
|
0.002
|
0.002
|
0.013
|
0.004
|
0.003
|
0.006
|
0.010
|
0.0004
|
0.0002
|
|
Adjusted R2
|
-0.006
|
-0.007
|
-0.008
|
-0.007
|
0.004
|
-0.006
|
-0.009
|
-0.004
|
0.0003
|
-0.001
|
-0.001
|
|
Residual Std. Error
|
0.265 (df = 435)
|
0.265 (df = 434)
|
0.265 (df = 434)
|
0.265 (df = 433)
|
0.254 (df = 414)
|
0.255 (df = 413)
|
0.256 (df = 412)
|
0.252 (df = 408)
|
0.255 (df = 411)
|
0.242 (df = 2717)
|
0.259 (df = 3797)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Gender book
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Gender book
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.042
|
0.061
|
0.068
|
0.022
|
0.042
|
0.034
|
0.055
|
0.042
|
0.014
|
0.037
|
0.032
|
|
|
(0.042)
|
(0.069)
|
(0.054)
|
(0.050)
|
(0.055)
|
(0.050)
|
(0.049)
|
(0.043)
|
(0.041)
|
(0.058)
|
(0.054)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
-0.190***
|
-0.020
|
0.045
|
-0.040
|
-0.011
|
0.123***
|
-0.112***
|
-0.067*
|
0.022
|
-0.043
|
-0.019
|
|
|
(0.030)
|
(0.086)
|
(0.052)
|
(0.089)
|
(0.078)
|
(0.032)
|
(0.029)
|
(0.039)
|
(0.036)
|
(0.041)
|
(0.034)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.063
|
-0.160**
|
-0.007
|
-0.052
|
|
-0.066
|
|
0.036
|
-0.020
|
-0.028
|
|
|
(0.000)
|
(0.109)
|
(0.077)
|
(0.098)
|
(0.091)
|
(0.000)
|
(0.055)
|
(0.000)
|
(0.060)
|
(0.054)
|
(0.049)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
439
|
439
|
439
|
438
|
419
|
418
|
418
|
413
|
416
|
2,721
|
3,801
|
|
R2
|
0.020
|
0.017
|
0.019
|
0.016
|
0.015
|
0.013
|
0.024
|
0.021
|
0.017
|
0.003
|
0.001
|
|
Adjusted R2
|
0.013
|
0.008
|
0.010
|
0.007
|
0.006
|
0.003
|
0.012
|
0.012
|
0.007
|
0.002
|
0.001
|
|
Residual Std. Error
|
0.491 (df = 435)
|
0.493 (df = 434)
|
0.492 (df = 434)
|
0.493 (df = 433)
|
0.492 (df = 414)
|
0.492 (df = 413)
|
0.491 (df = 412)
|
0.491 (df = 408)
|
0.492 (df = 411)
|
0.495 (df = 2717)
|
0.494 (df = 3797)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Morality book
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Morality book
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.048
|
-0.071
|
-0.035
|
-0.023
|
-0.066
|
-0.043
|
-0.065*
|
-0.060*
|
-0.0003
|
-0.059
|
-0.051
|
|
|
(0.034)
|
(0.053)
|
(0.028)
|
(0.036)
|
(0.047)
|
(0.036)
|
(0.037)
|
(0.033)
|
(0.021)
|
(0.044)
|
(0.041)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.094***
|
-0.071
|
0.015
|
0.048
|
-0.053
|
0.062*
|
-0.069**
|
0.070***
|
0.074
|
0.021
|
0.005
|
|
|
(0.025)
|
(0.043)
|
(0.046)
|
(0.084)
|
(0.051)
|
(0.032)
|
(0.030)
|
(0.026)
|
(0.058)
|
(0.033)
|
(0.027)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.111
|
0.015
|
-0.057
|
0.093*
|
|
0.169***
|
|
-0.150**
|
-0.001
|
0.013
|
|
|
(0.000)
|
(0.072)
|
(0.060)
|
(0.090)
|
(0.055)
|
(0.000)
|
(0.044)
|
(0.000)
|
(0.074)
|
(0.042)
|
(0.038)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
439
|
439
|
439
|
438
|
419
|
418
|
418
|
413
|
416
|
2,721
|
3,801
|
|
R2
|
0.017
|
0.018
|
0.015
|
0.016
|
0.014
|
0.011
|
0.016
|
0.019
|
0.022
|
0.006
|
0.003
|
|
Adjusted R2
|
0.010
|
0.009
|
0.006
|
0.007
|
0.004
|
0.002
|
0.004
|
0.009
|
0.012
|
0.005
|
0.003
|
|
Residual Std. Error
|
0.381 (df = 435)
|
0.381 (df = 434)
|
0.382 (df = 434)
|
0.382 (df = 433)
|
0.387 (df = 414)
|
0.390 (df = 413)
|
0.389 (df = 412)
|
0.384 (df = 408)
|
0.384 (df = 411)
|
0.380 (df = 2717)
|
0.385 (df = 3797)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|
Judicial independence book
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Judicial independence book
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.006
|
0.005
|
-0.033
|
-0.005
|
-0.020
|
-0.004
|
-0.007
|
0.016
|
0.023
|
0.030
|
0.019
|
|
|
(0.034)
|
(0.048)
|
(0.037)
|
(0.035)
|
(0.043)
|
(0.037)
|
(0.044)
|
(0.039)
|
(0.037)
|
(0.055)
|
(0.051)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.084**
|
0.085**
|
-0.064*
|
0.009
|
0.037
|
-0.267***
|
0.131***
|
0.076***
|
-0.020
|
0.023
|
0.016
|
|
|
(0.034)
|
(0.040)
|
(0.036)
|
(0.063)
|
(0.051)
|
(0.023)
|
(0.030)
|
(0.028)
|
(0.041)
|
(0.041)
|
(0.033)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.059
|
0.123**
|
0.075
|
0.072
|
|
-0.023
|
|
0.056
|
0.023
|
0.026
|
|
|
(0.000)
|
(0.064)
|
(0.050)
|
(0.070)
|
(0.062)
|
(0.000)
|
(0.060)
|
(0.000)
|
(0.052)
|
(0.053)
|
(0.048)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
439
|
439
|
439
|
438
|
419
|
418
|
418
|
413
|
416
|
2,721
|
3,801
|
|
R2
|
0.004
|
0.005
|
0.006
|
0.006
|
0.007
|
0.015
|
0.011
|
0.008
|
0.004
|
0.003
|
0.001
|
|
Adjusted R2
|
-0.003
|
-0.004
|
-0.003
|
-0.003
|
-0.003
|
0.005
|
-0.001
|
-0.002
|
-0.005
|
0.002
|
0.001
|
|
Residual Std. Error
|
0.470 (df = 435)
|
0.469 (df = 434)
|
0.469 (df = 434)
|
0.469 (df = 433)
|
0.471 (df = 414)
|
0.470 (df = 413)
|
0.470 (df = 412)
|
0.471 (df = 408)
|
0.472 (df = 411)
|
0.470 (df = 2717)
|
0.469 (df = 3797)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level.
|

Effect on exit survey answers
outcome_cols = c('work_satisfied_likert' = 'Satisfied with work',
'work_important_likert' = 'Work is important',
'motivated_likert' = 'Motivated to become better judge/prosecutor',
'trust_likert' = 'Trust in judicial system',
'learning_satisfied_likert' = 'Satisfaction with learning',
'improve_likert' = 'Training is key for improvement of judicial system',
'professor_likert' = 'Professors provide motivation',
'growth_1_likert' = 'Growth 1: I truly think people can change',
'growth_2_likert' = 'Growth 2: I can always change how smart I am',
"avg_growth_likert" = 'Average growth',
'challenge_1_likert' = 'Challenge 1: I enjoy projects that take years to be completed',
'challenge_2_likert' = 'Challenge 2: What I do everyday is connected to my deepest personal values',
'challenge_3_likert' = 'Challenge 3: There is at least one topic or activity that I never get bored thinking about',
'challenge_4_likert' = 'Challenge 4: I am hardworking',
"avg_challenge_likert" = 'Average challenge',
"empowerment_1_likert" = 'Empowerment 1: I question my own opinions, positions and points of view because they could be wrong',
"empowerment_2_likert" = "Empowerment 2: I recognize the value of opinions that are different than mine",
'avg_empowerment_likert' = 'Average empowerment',
"curiosity_1_likert" = 'Curiosity 1: I was so absorbed in learning that I lost track of time',
"curiosity_2_likert" = 'Curiosity 2: I explored a completely new idea or topic just for fun',
'avg_curiosity_likert' = 'Average curiosity',
"goal_1_likert" = 'Goal setting 1: I reflect on my life goals and the type of person I want to be',
"goal_2_likert" = 'Goal setting 2: When planning my day, I think about how my activities are related to what I want to achieve in life',
'avg_goal_likert' = 'Average goal setting',
"gratitude_1_likert" = 'Gratitude 1: I say "Thank you"',
"gratitude_2_likert" = 'Gratitude 2: I did something nice to show my appreciation',
"avg_gratitude_likert" = 'Average gratitude')
note <- "Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale."
for (outcome in names(outcome_cols)) {
formula <- as.formula(paste0(outcome, ' ~ monitoreo * teacher_female | idcurso | 0 | aula'))
out_list <- list()
for (r in 1:9) {
model <- felm(formula, data = data %>% filter(round == r))
out_list[[sprintf('Round %i', r)]] <- model
}
formula <- as.formula(paste0(outcome, ' ~ monitoreo * teacher_female | 0 | 0 | dni'))
out_list[['Pure Treatment']] <- felm(formula, data = pure_control)
out_list[['Perc. courses monitored']] <- felm(formula, data = monit_perc)
out_name <- outcome_cols[[outcome]]
cat('<h4></h4>')
cat(sprintf('<h3>%s</h3>', out_name))
stargazer(out_list,
covariate.labels = c('Monitoring', 'Female Teacher', 'Monitoring x Female Teacher'),
dep.var.labels = out_name,
column.labels = names(out_list),
notes = note,
omit = 'Constant',
type = 'html')
feols(fml = as.formula(sprintf('%s ~ i(round, monitoreo) | idcurso', outcome)),
data = data, cluster = ~aula) %>%
iplot(dict = outcome_cols, xlab = 'Round')
}
Satisfied with work
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Satisfied with work
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.015
|
-0.007
|
0.001
|
-0.035
|
0.030
|
0.029
|
0.061
|
0.046
|
0.009
|
0.021
|
0.021
|
|
|
(0.036)
|
(0.048)
|
(0.047)
|
(0.032)
|
(0.041)
|
(0.055)
|
(0.060)
|
(0.063)
|
(0.035)
|
(0.072)
|
(0.068)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
-0.019
|
0.095**
|
-0.036
|
-0.075
|
0.015
|
0.082*
|
0.210***
|
0.031
|
-0.049
|
0.009
|
-0.001
|
|
|
(0.026)
|
(0.039)
|
(0.051)
|
(0.059)
|
(0.069)
|
(0.045)
|
(0.042)
|
(0.053)
|
(0.035)
|
(0.047)
|
(0.039)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.038
|
-0.034
|
0.176**
|
-0.033
|
|
-0.215***
|
|
0.075
|
0.014
|
0.027
|
|
|
(0.000)
|
(0.086)
|
(0.064)
|
(0.077)
|
(0.079)
|
(0.000)
|
(0.068)
|
(0.000)
|
(0.072)
|
(0.065)
|
(0.061)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
459
|
459
|
459
|
458
|
439
|
438
|
438
|
433
|
436
|
2,864
|
3,978
|
|
R2
|
0.002
|
0.004
|
0.003
|
0.005
|
0.001
|
0.002
|
0.006
|
0.003
|
0.001
|
0.001
|
0.0005
|
|
Adjusted R2
|
-0.005
|
-0.005
|
-0.006
|
-0.004
|
-0.008
|
-0.008
|
-0.006
|
-0.006
|
-0.008
|
-0.001
|
-0.0003
|
|
Residual Std. Error
|
0.658 (df = 455)
|
0.658 (df = 454)
|
0.658 (df = 454)
|
0.658 (df = 453)
|
0.665 (df = 434)
|
0.666 (df = 433)
|
0.665 (df = 432)
|
0.660 (df = 428)
|
0.667 (df = 431)
|
0.631 (df = 2860)
|
0.659 (df = 3974)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Work is important
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Work is important
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.014
|
-0.026
|
-0.006
|
-0.046
|
0.030
|
0.024
|
0.059
|
0.044
|
0.024
|
0.023
|
0.019
|
|
|
(0.047)
|
(0.061)
|
(0.054)
|
(0.041)
|
(0.060)
|
(0.061)
|
(0.063)
|
(0.066)
|
(0.033)
|
(0.074)
|
(0.070)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.046
|
0.098*
|
-0.050
|
-0.106
|
0.023
|
0.0005
|
0.216***
|
0.078
|
-0.119***
|
-0.007
|
-0.013
|
|
|
(0.041)
|
(0.051)
|
(0.060)
|
(0.069)
|
(0.076)
|
(0.049)
|
(0.046)
|
(0.058)
|
(0.035)
|
(0.048)
|
(0.040)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.015
|
0.012
|
0.242**
|
-0.002
|
|
-0.174**
|
|
0.129*
|
0.049
|
0.066
|
|
|
(0.000)
|
(0.102)
|
(0.083)
|
(0.098)
|
(0.090)
|
(0.000)
|
(0.069)
|
(0.000)
|
(0.075)
|
(0.066)
|
(0.062)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
459
|
459
|
459
|
458
|
439
|
438
|
438
|
433
|
436
|
2,864
|
3,978
|
|
R2
|
0.002
|
0.006
|
0.002
|
0.008
|
0.001
|
0.001
|
0.006
|
0.003
|
0.004
|
0.001
|
0.001
|
|
Adjusted R2
|
-0.005
|
-0.002
|
-0.006
|
-0.001
|
-0.008
|
-0.009
|
-0.005
|
-0.006
|
-0.005
|
0.0001
|
0.0003
|
|
Residual Std. Error
|
0.669 (df = 455)
|
0.669 (df = 454)
|
0.670 (df = 454)
|
0.669 (df = 453)
|
0.678 (df = 434)
|
0.679 (df = 433)
|
0.678 (df = 432)
|
0.674 (df = 428)
|
0.678 (df = 431)
|
0.648 (df = 2860)
|
0.671 (df = 3974)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Motivated to become better judge/prosecutor
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Motivated to become better judge/prosecutor
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.006
|
-0.016
|
-0.038
|
-0.054
|
0.022
|
0.007
|
0.049
|
0.030
|
0.015
|
0.008
|
0.006
|
|
|
(0.045)
|
(0.055)
|
(0.054)
|
(0.037)
|
(0.051)
|
(0.059)
|
(0.062)
|
(0.064)
|
(0.038)
|
(0.074)
|
(0.071)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.027
|
0.085**
|
-0.077
|
-0.114
|
0.007
|
-0.079
|
0.200***
|
0.069
|
-0.089**
|
-0.016
|
-0.024
|
|
|
(0.049)
|
(0.042)
|
(0.058)
|
(0.071)
|
(0.080)
|
(0.048)
|
(0.047)
|
(0.054)
|
(0.040)
|
(0.048)
|
(0.040)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.010
|
0.077
|
0.225**
|
-0.004
|
|
-0.173**
|
|
0.075
|
0.051
|
0.064
|
|
|
(0.000)
|
(0.092)
|
(0.068)
|
(0.098)
|
(0.093)
|
(0.000)
|
(0.068)
|
(0.000)
|
(0.073)
|
(0.067)
|
(0.062)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
459
|
459
|
459
|
458
|
439
|
438
|
438
|
433
|
436
|
2,864
|
3,978
|
|
R2
|
0.001
|
0.003
|
0.002
|
0.006
|
0.0005
|
0.001
|
0.004
|
0.003
|
0.002
|
0.001
|
0.001
|
|
Adjusted R2
|
-0.006
|
-0.005
|
-0.006
|
-0.003
|
-0.009
|
-0.009
|
-0.007
|
-0.007
|
-0.008
|
-0.001
|
-0.0002
|
|
Residual Std. Error
|
0.676 (df = 455)
|
0.676 (df = 454)
|
0.676 (df = 454)
|
0.675 (df = 453)
|
0.684 (df = 434)
|
0.685 (df = 433)
|
0.684 (df = 432)
|
0.678 (df = 428)
|
0.686 (df = 431)
|
0.651 (df = 2860)
|
0.677 (df = 3974)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Trust in judicial system
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Trust in judicial system
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.017
|
-0.036
|
-0.044
|
-0.064**
|
-0.016
|
-0.009
|
0.030
|
0.004
|
-0.043**
|
-0.035
|
-0.021
|
|
|
(0.040)
|
(0.050)
|
(0.039)
|
(0.027)
|
(0.053)
|
(0.062)
|
(0.065)
|
(0.071)
|
(0.021)
|
(0.068)
|
(0.065)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.038
|
0.100***
|
-0.057
|
-0.081
|
0.012
|
-0.027
|
0.255***
|
0.054
|
-0.086
|
-0.018
|
-0.016
|
|
|
(0.027)
|
(0.036)
|
(0.051)
|
(0.083)
|
(0.064)
|
(0.051)
|
(0.047)
|
(0.059)
|
(0.090)
|
(0.047)
|
(0.039)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.027
|
0.050
|
0.210**
|
-0.006
|
|
-0.226***
|
|
0.169*
|
0.068
|
0.070
|
|
|
(0.000)
|
(0.090)
|
(0.060)
|
(0.095)
|
(0.073)
|
(0.000)
|
(0.071)
|
(0.000)
|
(0.102)
|
(0.065)
|
(0.060)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
458
|
458
|
458
|
457
|
438
|
437
|
437
|
432
|
435
|
2,864
|
3,969
|
|
R2
|
0.003
|
0.007
|
0.004
|
0.008
|
0.003
|
0.003
|
0.010
|
0.005
|
0.004
|
0.001
|
0.001
|
|
Adjusted R2
|
-0.003
|
-0.002
|
-0.005
|
-0.0004
|
-0.006
|
-0.006
|
-0.002
|
-0.004
|
-0.005
|
-0.0001
|
-0.0002
|
|
Residual Std. Error
|
0.637 (df = 454)
|
0.636 (df = 453)
|
0.637 (df = 453)
|
0.636 (df = 452)
|
0.643 (df = 433)
|
0.643 (df = 432)
|
0.641 (df = 431)
|
0.639 (df = 427)
|
0.643 (df = 430)
|
0.609 (df = 2860)
|
0.637 (df = 3965)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Satisfaction with learning
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Satisfaction with learning
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.044
|
0.033
|
0.041
|
0.005
|
0.075
|
0.064
|
0.095
|
0.069
|
0.069**
|
0.060
|
0.075
|
|
|
(0.039)
|
(0.057)
|
(0.050)
|
(0.038)
|
(0.047)
|
(0.052)
|
(0.059)
|
(0.057)
|
(0.033)
|
(0.074)
|
(0.070)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.027
|
0.078*
|
-0.018
|
-0.057
|
0.047
|
0.043
|
0.165***
|
0.082*
|
-0.087***
|
0.002
|
0.0004
|
|
|
(0.034)
|
(0.040)
|
(0.050)
|
(0.055)
|
(0.058)
|
(0.042)
|
(0.043)
|
(0.048)
|
(0.024)
|
(0.050)
|
(0.042)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.039
|
-0.004
|
0.171**
|
-0.032
|
|
-0.150**
|
|
0.068
|
0.020
|
0.027
|
|
|
(0.000)
|
(0.085)
|
(0.066)
|
(0.067)
|
(0.073)
|
(0.000)
|
(0.072)
|
(0.000)
|
(0.056)
|
(0.069)
|
(0.064)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
454
|
454
|
454
|
453
|
435
|
434
|
434
|
429
|
432
|
2,847
|
3,939
|
|
R2
|
0.001
|
0.002
|
0.002
|
0.005
|
0.003
|
0.003
|
0.007
|
0.007
|
0.005
|
0.003
|
0.003
|
|
Adjusted R2
|
-0.005
|
-0.007
|
-0.007
|
-0.004
|
-0.006
|
-0.006
|
-0.005
|
-0.003
|
-0.005
|
0.002
|
0.002
|
|
Residual Std. Error
|
0.678 (df = 450)
|
0.678 (df = 449)
|
0.679 (df = 449)
|
0.678 (df = 448)
|
0.678 (df = 430)
|
0.678 (df = 429)
|
0.678 (df = 428)
|
0.672 (df = 424)
|
0.682 (df = 427)
|
0.650 (df = 2843)
|
0.675 (df = 3935)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Training is key for improvement of judicial system
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Training is key for improvement of judicial system
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.008
|
-0.022
|
-0.037
|
-0.047
|
0.032
|
0.020
|
0.062
|
0.042
|
0.025
|
0.012
|
0.015
|
|
|
(0.044)
|
(0.056)
|
(0.057)
|
(0.041)
|
(0.050)
|
(0.057)
|
(0.060)
|
(0.061)
|
(0.037)
|
(0.074)
|
(0.071)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.034
|
0.083
|
-0.081
|
-0.127*
|
0.038
|
-0.044
|
0.200***
|
0.035
|
-0.093**
|
-0.012
|
-0.017
|
|
|
(0.034)
|
(0.055)
|
(0.064)
|
(0.070)
|
(0.090)
|
(0.049)
|
(0.049)
|
(0.050)
|
(0.046)
|
(0.047)
|
(0.039)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.011
|
0.083
|
0.226***
|
-0.019
|
|
-0.189***
|
|
0.064
|
0.042
|
0.054
|
|
|
(0.000)
|
(0.090)
|
(0.073)
|
(0.085)
|
(0.096)
|
(0.000)
|
(0.070)
|
(0.000)
|
(0.082)
|
(0.066)
|
(0.062)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
460
|
460
|
460
|
459
|
439
|
438
|
438
|
433
|
436
|
2,864
|
3,982
|
|
R2
|
0.001
|
0.004
|
0.002
|
0.006
|
0.001
|
0.0004
|
0.005
|
0.002
|
0.002
|
0.001
|
0.001
|
|
Adjusted R2
|
-0.006
|
-0.005
|
-0.006
|
-0.003
|
-0.008
|
-0.009
|
-0.007
|
-0.007
|
-0.007
|
-0.001
|
-0.0002
|
|
Residual Std. Error
|
0.675 (df = 456)
|
0.675 (df = 455)
|
0.675 (df = 455)
|
0.674 (df = 454)
|
0.684 (df = 434)
|
0.684 (df = 433)
|
0.684 (df = 432)
|
0.678 (df = 428)
|
0.685 (df = 431)
|
0.650 (df = 2860)
|
0.677 (df = 3978)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Professors provide motivation
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Professors provide motivation
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.059
|
0.029
|
0.077*
|
0.016
|
0.091
|
0.081
|
0.134**
|
0.105
|
0.052
|
0.095
|
0.095
|
|
|
(0.043)
|
(0.057)
|
(0.045)
|
(0.043)
|
(0.061)
|
(0.057)
|
(0.064)
|
(0.065)
|
(0.051)
|
(0.073)
|
(0.069)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
-0.088***
|
0.116***
|
0.021
|
-0.044
|
0.041
|
0.063
|
0.218***
|
0.017
|
-0.129***
|
0.025
|
0.025
|
|
|
(0.032)
|
(0.042)
|
(0.059)
|
(0.078)
|
(0.065)
|
(0.049)
|
(0.046)
|
(0.061)
|
(0.030)
|
(0.049)
|
(0.041)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.066
|
-0.102
|
0.148
|
-0.058
|
|
-0.275***
|
|
0.166**
|
-0.023
|
-0.017
|
|
|
(0.000)
|
(0.092)
|
(0.063)
|
(0.091)
|
(0.073)
|
(0.000)
|
(0.086)
|
(0.000)
|
(0.077)
|
(0.069)
|
(0.064)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
459
|
459
|
459
|
458
|
439
|
438
|
438
|
433
|
436
|
2,864
|
3,978
|
|
R2
|
0.002
|
0.004
|
0.003
|
0.005
|
0.004
|
0.004
|
0.011
|
0.007
|
0.009
|
0.005
|
0.004
|
|
Adjusted R2
|
-0.004
|
-0.005
|
-0.006
|
-0.004
|
-0.006
|
-0.006
|
-0.001
|
-0.002
|
-0.0005
|
0.004
|
0.003
|
|
Residual Std. Error
|
0.667 (df = 455)
|
0.667 (df = 454)
|
0.667 (df = 454)
|
0.667 (df = 453)
|
0.668 (df = 434)
|
0.668 (df = 433)
|
0.667 (df = 432)
|
0.664 (df = 428)
|
0.668 (df = 431)
|
0.646 (df = 2860)
|
0.664 (df = 3974)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Growth 1: I truly think people can change
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Growth 1: I truly think people can change
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.069*
|
-0.076
|
-0.052
|
-0.086*
|
-0.074
|
-0.052
|
-0.042
|
-0.050
|
-0.005
|
-0.078
|
-0.061
|
|
|
(0.039)
|
(0.062)
|
(0.046)
|
(0.044)
|
(0.046)
|
(0.042)
|
(0.044)
|
(0.044)
|
(0.032)
|
(0.051)
|
(0.049)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.078**
|
-0.005
|
0.033
|
-0.029
|
-0.074**
|
0.058*
|
0.151***
|
0.054
|
0.013
|
0.018
|
0.026
|
|
|
(0.030)
|
(0.030)
|
(0.034)
|
(0.048)
|
(0.035)
|
(0.034)
|
(0.027)
|
(0.045)
|
(0.065)
|
(0.036)
|
(0.030)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.010
|
0.028
|
0.107
|
0.060
|
|
-0.080
|
|
-0.010
|
0.007
|
-0.002
|
|
|
(0.000)
|
(0.081)
|
(0.048)
|
(0.069)
|
(0.077)
|
(0.000)
|
(0.049)
|
(0.000)
|
(0.079)
|
(0.052)
|
(0.049)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
453
|
453
|
453
|
452
|
433
|
432
|
432
|
427
|
430
|
2,847
|
3,926
|
|
R2
|
0.007
|
0.008
|
0.008
|
0.009
|
0.012
|
0.010
|
0.013
|
0.008
|
0.011
|
0.007
|
0.004
|
|
Adjusted R2
|
0.001
|
-0.001
|
-0.001
|
0.0004
|
0.002
|
0.0003
|
0.002
|
-0.001
|
0.002
|
0.006
|
0.003
|
|
Residual Std. Error
|
0.461 (df = 449)
|
0.462 (df = 448)
|
0.462 (df = 448)
|
0.462 (df = 447)
|
0.460 (df = 428)
|
0.461 (df = 427)
|
0.461 (df = 426)
|
0.457 (df = 422)
|
0.467 (df = 425)
|
0.452 (df = 2843)
|
0.462 (df = 3922)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Growth 2: I can always change how smart I am
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Growth 2: I can always change how smart I am
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.085*
|
-0.052
|
-0.116**
|
-0.060
|
-0.099*
|
-0.048
|
-0.049
|
-0.059
|
0.060
|
-0.044
|
-0.049
|
|
|
(0.047)
|
(0.081)
|
(0.059)
|
(0.054)
|
(0.059)
|
(0.054)
|
(0.055)
|
(0.057)
|
(0.040)
|
(0.059)
|
(0.056)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.150***
|
0.065
|
-0.032
|
0.015
|
-0.086
|
0.085**
|
0.195***
|
0.104*
|
0.136*
|
0.044
|
0.056*
|
|
|
(0.036)
|
(0.053)
|
(0.058)
|
(0.082)
|
(0.065)
|
(0.043)
|
(0.033)
|
(0.058)
|
(0.075)
|
(0.039)
|
(0.032)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.070
|
0.185***
|
-0.058
|
0.048
|
|
-0.103*
|
|
-0.211**
|
-0.038
|
-0.056
|
|
|
(0.000)
|
(0.097)
|
(0.068)
|
(0.092)
|
(0.080)
|
(0.000)
|
(0.062)
|
(0.000)
|
(0.093)
|
(0.055)
|
(0.052)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
445
|
445
|
445
|
444
|
425
|
425
|
424
|
419
|
422
|
2,793
|
3,855
|
|
R2
|
0.016
|
0.014
|
0.019
|
0.014
|
0.016
|
0.015
|
0.022
|
0.016
|
0.020
|
0.003
|
0.003
|
|
Adjusted R2
|
0.009
|
0.005
|
0.010
|
0.005
|
0.006
|
0.005
|
0.010
|
0.006
|
0.011
|
0.002
|
0.003
|
|
Residual Std. Error
|
0.529 (df = 441)
|
0.530 (df = 440)
|
0.529 (df = 440)
|
0.530 (df = 439)
|
0.525 (df = 420)
|
0.527 (df = 420)
|
0.526 (df = 418)
|
0.523 (df = 414)
|
0.530 (df = 417)
|
0.514 (df = 2789)
|
0.529 (df = 3851)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Average growth
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Average growth
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.077*
|
-0.066
|
-0.086*
|
-0.074
|
-0.087*
|
-0.053
|
-0.046
|
-0.057
|
0.022
|
-0.063
|
-0.056
|
|
|
(0.040)
|
(0.066)
|
(0.049)
|
(0.046)
|
(0.048)
|
(0.045)
|
(0.046)
|
(0.049)
|
(0.034)
|
(0.050)
|
(0.047)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.092***
|
0.029
|
-0.001
|
-0.008
|
-0.080*
|
0.071*
|
0.172***
|
0.072
|
0.074
|
0.030
|
0.042
|
|
|
(0.029)
|
(0.039)
|
(0.044)
|
(0.064)
|
(0.048)
|
(0.036)
|
(0.028)
|
(0.050)
|
(0.070)
|
(0.034)
|
(0.028)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.031
|
0.099*
|
0.020
|
0.046
|
|
-0.106**
|
|
-0.103
|
-0.020
|
-0.034
|
|
|
(0.000)
|
(0.080)
|
(0.054)
|
(0.075)
|
(0.068)
|
(0.000)
|
(0.051)
|
(0.000)
|
(0.083)
|
(0.050)
|
(0.047)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
445
|
445
|
445
|
444
|
425
|
425
|
424
|
419
|
422
|
2,793
|
3,855
|
|
R2
|
0.014
|
0.013
|
0.015
|
0.012
|
0.018
|
0.015
|
0.021
|
0.015
|
0.018
|
0.006
|
0.005
|
|
Adjusted R2
|
0.007
|
0.004
|
0.006
|
0.003
|
0.008
|
0.005
|
0.009
|
0.005
|
0.008
|
0.005
|
0.004
|
|
Residual Std. Error
|
0.450 (df = 441)
|
0.451 (df = 440)
|
0.451 (df = 440)
|
0.452 (df = 439)
|
0.450 (df = 420)
|
0.451 (df = 420)
|
0.450 (df = 418)
|
0.445 (df = 414)
|
0.456 (df = 417)
|
0.439 (df = 2789)
|
0.452 (df = 3851)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Challenge 1: I enjoy projects that take years to be completed
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Challenge 1: I enjoy projects that take years to be completed
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.047
|
0.017
|
-0.121
|
-0.018
|
-0.063
|
-0.041
|
-0.077
|
-0.093
|
0.055
|
-0.065
|
-0.072
|
|
|
(0.069)
|
(0.094)
|
(0.099)
|
(0.062)
|
(0.084)
|
(0.068)
|
(0.068)
|
(0.066)
|
(0.053)
|
(0.069)
|
(0.065)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.084
|
-0.053
|
-0.099
|
-0.160***
|
-0.097
|
-0.081
|
-0.117**
|
0.175***
|
0.027
|
-0.103**
|
-0.091**
|
|
|
(0.057)
|
(0.074)
|
(0.082)
|
(0.058)
|
(0.086)
|
(0.051)
|
(0.051)
|
(0.061)
|
(0.207)
|
(0.048)
|
(0.040)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.072
|
0.259**
|
-0.101
|
0.043
|
|
0.287***
|
|
-0.211
|
0.069
|
0.074
|
|
|
(0.000)
|
(0.154)
|
(0.109)
|
(0.137)
|
(0.140)
|
(0.000)
|
(0.075)
|
(0.000)
|
(0.217)
|
(0.067)
|
(0.063)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
451
|
451
|
451
|
450
|
431
|
430
|
430
|
426
|
429
|
2,838
|
3,910
|
|
R2
|
0.002
|
0.006
|
0.009
|
0.022
|
0.006
|
0.004
|
0.009
|
0.008
|
0.012
|
0.004
|
0.003
|
|
Adjusted R2
|
-0.004
|
-0.003
|
0.0001
|
0.013
|
-0.004
|
-0.005
|
-0.003
|
-0.002
|
0.002
|
0.003
|
0.003
|
|
Residual Std. Error
|
0.636 (df = 447)
|
0.637 (df = 446)
|
0.636 (df = 446)
|
0.632 (df = 445)
|
0.637 (df = 426)
|
0.636 (df = 425)
|
0.634 (df = 424)
|
0.634 (df = 421)
|
0.631 (df = 424)
|
0.612 (df = 2834)
|
0.633 (df = 3906)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Challenge 2: What I do everyday is connected to my deepest personal values
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Challenge 2: What I do everyday is connected to my deepest personal values
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.050
|
-0.077
|
-0.062
|
-0.071
|
-0.075
|
-0.029
|
-0.030
|
-0.028
|
0.068**
|
-0.037
|
-0.036
|
|
|
(0.038)
|
(0.057)
|
(0.045)
|
(0.046)
|
(0.050)
|
(0.047)
|
(0.052)
|
(0.045)
|
(0.030)
|
(0.049)
|
(0.047)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.077*
|
0.034
|
-0.022
|
-0.065
|
-0.051
|
0.017
|
0.149***
|
0.038
|
0.079
|
-0.001
|
0.016
|
|
|
(0.041)
|
(0.062)
|
(0.041)
|
(0.042)
|
(0.044)
|
(0.036)
|
(0.032)
|
(0.045)
|
(0.063)
|
(0.032)
|
(0.027)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.030
|
0.047
|
0.107*
|
0.074
|
|
-0.120**
|
|
-0.118
|
0.006
|
0.0002
|
|
|
(0.000)
|
(0.087)
|
(0.061)
|
(0.055)
|
(0.062)
|
(0.000)
|
(0.055)
|
(0.000)
|
(0.082)
|
(0.051)
|
(0.048)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
450
|
450
|
450
|
449
|
430
|
429
|
429
|
425
|
427
|
2,838
|
3,900
|
|
R2
|
0.004
|
0.007
|
0.003
|
0.006
|
0.005
|
0.006
|
0.011
|
0.005
|
0.007
|
0.002
|
0.001
|
|
Adjusted R2
|
-0.003
|
-0.002
|
-0.005
|
-0.003
|
-0.004
|
-0.003
|
-0.001
|
-0.005
|
-0.003
|
0.001
|
0.001
|
|
Residual Std. Error
|
0.454 (df = 446)
|
0.453 (df = 445)
|
0.454 (df = 445)
|
0.454 (df = 444)
|
0.453 (df = 425)
|
0.454 (df = 424)
|
0.453 (df = 423)
|
0.448 (df = 420)
|
0.462 (df = 422)
|
0.446 (df = 2834)
|
0.454 (df = 3896)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Challenge 3: There is at least one topic or activity that I never get bored thinking about
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Challenge 3: There is at least one topic or activity that I never get bored thinking about
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.074*
|
-0.099*
|
-0.110**
|
-0.073*
|
-0.106**
|
-0.125***
|
-0.121***
|
-0.138***
|
-0.024
|
-0.128**
|
-0.109*
|
|
|
(0.039)
|
(0.056)
|
(0.047)
|
(0.043)
|
(0.053)
|
(0.040)
|
(0.037)
|
(0.040)
|
(0.043)
|
(0.060)
|
(0.056)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.008
|
0.011
|
-0.070
|
0.019
|
-0.062
|
-0.074*
|
0.175***
|
0.074**
|
0.174**
|
-0.016
|
-0.007
|
|
|
(0.033)
|
(0.046)
|
(0.057)
|
(0.074)
|
(0.063)
|
(0.042)
|
(0.035)
|
(0.030)
|
(0.081)
|
(0.035)
|
(0.029)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.041
|
0.043
|
-0.006
|
0.067
|
|
-0.127***
|
|
-0.168*
|
0.029
|
0.023
|
|
|
(0.000)
|
(0.071)
|
(0.063)
|
(0.078)
|
(0.065)
|
(0.000)
|
(0.045)
|
(0.000)
|
(0.086)
|
(0.057)
|
(0.053)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
452
|
452
|
452
|
451
|
432
|
431
|
431
|
426
|
429
|
2,838
|
3,917
|
|
R2
|
0.011
|
0.013
|
0.014
|
0.011
|
0.013
|
0.031
|
0.036
|
0.035
|
0.012
|
0.014
|
0.008
|
|
Adjusted R2
|
0.004
|
0.004
|
0.005
|
0.002
|
0.003
|
0.021
|
0.025
|
0.026
|
0.003
|
0.012
|
0.007
|
|
Residual Std. Error
|
0.524 (df = 448)
|
0.524 (df = 447)
|
0.524 (df = 447)
|
0.525 (df = 446)
|
0.525 (df = 427)
|
0.521 (df = 426)
|
0.520 (df = 425)
|
0.519 (df = 421)
|
0.530 (df = 424)
|
0.525 (df = 2834)
|
0.525 (df = 3913)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Challenge 4: I am hardworking
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Challenge 4: I am hardworking
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.039
|
-0.103*
|
-0.085
|
-0.074
|
-0.052
|
-0.037
|
-0.034
|
-0.058
|
0.021
|
-0.051
|
-0.052
|
|
|
(0.048)
|
(0.060)
|
(0.053)
|
(0.049)
|
(0.064)
|
(0.052)
|
(0.057)
|
(0.058)
|
(0.050)
|
(0.049)
|
(0.047)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.053
|
-0.047
|
-0.085*
|
0.010
|
-0.041
|
-0.093***
|
0.122***
|
0.062
|
0.111*
|
-0.004
|
-0.008
|
|
|
(0.058)
|
(0.035)
|
(0.049)
|
(0.068)
|
(0.052)
|
(0.034)
|
(0.032)
|
(0.059)
|
(0.067)
|
(0.031)
|
(0.026)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.172*
|
0.101
|
0.158*
|
0.055
|
|
-0.068
|
|
-0.100
|
0.049
|
0.064
|
|
|
(0.000)
|
(0.102)
|
(0.071)
|
(0.091)
|
(0.080)
|
(0.000)
|
(0.060)
|
(0.000)
|
(0.084)
|
(0.050)
|
(0.047)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
449
|
449
|
449
|
448
|
429
|
428
|
428
|
423
|
426
|
2,838
|
3,890
|
|
R2
|
0.002
|
0.010
|
0.006
|
0.014
|
0.003
|
0.005
|
0.008
|
0.007
|
0.005
|
0.003
|
0.003
|
|
Adjusted R2
|
-0.005
|
0.001
|
-0.003
|
0.006
|
-0.007
|
-0.004
|
-0.004
|
-0.003
|
-0.004
|
0.002
|
0.002
|
|
Residual Std. Error
|
0.447 (df = 445)
|
0.446 (df = 444)
|
0.447 (df = 444)
|
0.445 (df = 443)
|
0.454 (df = 424)
|
0.454 (df = 423)
|
0.454 (df = 422)
|
0.448 (df = 418)
|
0.454 (df = 421)
|
0.438 (df = 2834)
|
0.450 (df = 3886)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Average challenge
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Average challenge
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.055
|
-0.069
|
-0.099**
|
-0.068
|
-0.083*
|
-0.062
|
-0.066
|
-0.080**
|
0.020
|
-0.070
|
-0.072*
|
|
|
(0.036)
|
(0.054)
|
(0.048)
|
(0.043)
|
(0.048)
|
(0.040)
|
(0.041)
|
(0.039)
|
(0.033)
|
(0.043)
|
(0.041)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.037
|
-0.015
|
-0.070
|
-0.050
|
-0.064
|
-0.096***
|
0.082***
|
0.095***
|
0.096
|
-0.031
|
-0.023
|
|
|
(0.035)
|
(0.042)
|
(0.046)
|
(0.049)
|
(0.048)
|
(0.032)
|
(0.030)
|
(0.036)
|
(0.101)
|
(0.027)
|
(0.023)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.049
|
0.106*
|
0.056
|
0.073
|
|
-0.009
|
|
-0.141
|
0.042
|
0.044
|
|
|
(0.000)
|
(0.067)
|
(0.055)
|
(0.058)
|
(0.057)
|
(0.000)
|
(0.044)
|
(0.000)
|
(0.106)
|
(0.045)
|
(0.042)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
443
|
443
|
443
|
442
|
423
|
422
|
422
|
419
|
421
|
2,811
|
3,839
|
|
R2
|
0.005
|
0.006
|
0.010
|
0.007
|
0.009
|
0.010
|
0.012
|
0.014
|
0.007
|
0.007
|
0.006
|
|
Adjusted R2
|
-0.001
|
-0.003
|
0.001
|
-0.002
|
-0.001
|
0.0002
|
-0.0001
|
0.004
|
-0.003
|
0.006
|
0.005
|
|
Residual Std. Error
|
0.396 (df = 439)
|
0.396 (df = 438)
|
0.395 (df = 438)
|
0.396 (df = 437)
|
0.398 (df = 418)
|
0.398 (df = 417)
|
0.398 (df = 416)
|
0.393 (df = 414)
|
0.400 (df = 416)
|
0.383 (df = 2807)
|
0.396 (df = 3835)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Empowerment 1: I question my own opinions, positions and points of view because they could be wrong
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Empowerment 1: I question my own opinions, positions and points of view because they could be wrong
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.028
|
0.033
|
0.034
|
0.024
|
0.015
|
0.055
|
0.024
|
0.021
|
0.130***
|
0.048
|
0.045
|
|
|
(0.043)
|
(0.056)
|
(0.052)
|
(0.047)
|
(0.052)
|
(0.046)
|
(0.045)
|
(0.042)
|
(0.040)
|
(0.066)
|
(0.063)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.134***
|
-0.008
|
0.008
|
-0.087
|
-0.106*
|
-0.028
|
-0.021
|
-0.006
|
0.085**
|
-0.070
|
-0.055
|
|
|
(0.045)
|
(0.092)
|
(0.051)
|
(0.074)
|
(0.059)
|
(0.034)
|
(0.034)
|
(0.033)
|
(0.037)
|
(0.053)
|
(0.043)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.021
|
0.044
|
0.052
|
0.098
|
|
0.034
|
|
-0.168**
|
0.079
|
0.060
|
|
|
(0.000)
|
(0.108)
|
(0.079)
|
(0.085)
|
(0.071)
|
(0.000)
|
(0.080)
|
(0.000)
|
(0.069)
|
(0.069)
|
(0.063)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
448
|
448
|
448
|
447
|
428
|
427
|
427
|
422
|
425
|
2,820
|
3,881
|
|
R2
|
0.009
|
0.006
|
0.007
|
0.009
|
0.011
|
0.025
|
0.022
|
0.018
|
0.011
|
0.004
|
0.002
|
|
Adjusted R2
|
0.002
|
-0.002
|
-0.002
|
0.00005
|
0.002
|
0.016
|
0.011
|
0.008
|
0.002
|
0.003
|
0.002
|
|
Residual Std. Error
|
0.607 (df = 444)
|
0.609 (df = 443)
|
0.608 (df = 443)
|
0.609 (df = 442)
|
0.608 (df = 423)
|
0.601 (df = 422)
|
0.605 (df = 421)
|
0.608 (df = 417)
|
0.606 (df = 420)
|
0.595 (df = 2816)
|
0.607 (df = 3877)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Empowerment 2: I recognize the value of opinions that are different than mine
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Empowerment 2: I recognize the value of opinions that are different than mine
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.028
|
-0.063
|
-0.066
|
-0.051
|
-0.042
|
-0.023
|
-0.014
|
-0.016
|
0.082**
|
-0.004
|
-0.020
|
|
|
(0.045)
|
(0.064)
|
(0.054)
|
(0.049)
|
(0.055)
|
(0.050)
|
(0.054)
|
(0.050)
|
(0.041)
|
(0.049)
|
(0.046)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.006
|
-0.025
|
-0.067
|
-0.092
|
-0.089
|
0.010
|
0.068*
|
-0.050
|
0.124*
|
-0.025
|
-0.021
|
|
|
(0.046)
|
(0.077)
|
(0.057)
|
(0.081)
|
(0.063)
|
(0.038)
|
(0.038)
|
(0.046)
|
(0.065)
|
(0.032)
|
(0.026)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.070
|
0.083
|
0.093
|
0.065
|
|
-0.080
|
|
-0.150*
|
0.013
|
0.019
|
|
|
(0.000)
|
(0.101)
|
(0.067)
|
(0.093)
|
(0.078)
|
(0.000)
|
(0.063)
|
(0.000)
|
(0.079)
|
(0.050)
|
(0.046)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
450
|
450
|
450
|
449
|
429
|
428
|
428
|
423
|
426
|
2,820
|
3,894
|
|
R2
|
0.002
|
0.004
|
0.005
|
0.006
|
0.005
|
0.010
|
0.011
|
0.012
|
0.009
|
0.0003
|
0.0004
|
|
Adjusted R2
|
-0.005
|
-0.005
|
-0.004
|
-0.003
|
-0.005
|
0.001
|
-0.001
|
0.003
|
-0.001
|
-0.001
|
-0.0003
|
|
Residual Std. Error
|
0.440 (df = 446)
|
0.440 (df = 445)
|
0.439 (df = 445)
|
0.439 (df = 444)
|
0.446 (df = 424)
|
0.445 (df = 423)
|
0.446 (df = 422)
|
0.439 (df = 418)
|
0.446 (df = 421)
|
0.435 (df = 2816)
|
0.443 (df = 3890)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Average empowerment
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Average empowerment
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.001
|
-0.013
|
-0.016
|
-0.008
|
-0.009
|
0.019
|
0.005
|
0.006
|
0.113***
|
0.026
|
0.015
|
|
|
(0.040)
|
(0.059)
|
(0.049)
|
(0.044)
|
(0.052)
|
(0.045)
|
(0.047)
|
(0.045)
|
(0.034)
|
(0.051)
|
(0.047)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.116***
|
-0.015
|
-0.029
|
-0.087
|
-0.095
|
-0.007
|
0.026
|
-0.029
|
0.107**
|
-0.048
|
-0.038
|
|
|
(0.033)
|
(0.084)
|
(0.051)
|
(0.077)
|
(0.058)
|
(0.034)
|
(0.035)
|
(0.039)
|
(0.049)
|
(0.037)
|
(0.029)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.027
|
0.085
|
0.065
|
0.077
|
|
-0.004
|
|
-0.167**
|
0.052
|
0.046
|
|
|
(0.000)
|
(0.103)
|
(0.069)
|
(0.085)
|
(0.071)
|
(0.000)
|
(0.076)
|
(0.000)
|
(0.068)
|
(0.053)
|
(0.048)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
446
|
446
|
446
|
445
|
426
|
425
|
425
|
420
|
423
|
2,802
|
3,863
|
|
R2
|
0.008
|
0.006
|
0.007
|
0.009
|
0.010
|
0.022
|
0.022
|
0.021
|
0.014
|
0.002
|
0.001
|
|
Adjusted R2
|
0.001
|
-0.003
|
-0.002
|
0.0004
|
0.0003
|
0.013
|
0.010
|
0.011
|
0.004
|
0.001
|
0.0003
|
|
Residual Std. Error
|
0.450 (df = 442)
|
0.451 (df = 441)
|
0.451 (df = 441)
|
0.451 (df = 440)
|
0.454 (df = 421)
|
0.451 (df = 420)
|
0.452 (df = 419)
|
0.451 (df = 415)
|
0.452 (df = 418)
|
0.451 (df = 2798)
|
0.452 (df = 3859)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Curiosity 1: I was so absorbed in learning that I lost track of time
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Curiosity 1: I was so absorbed in learning that I lost track of time
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.097*
|
-0.094
|
-0.127*
|
-0.085
|
0.010
|
-0.029
|
-0.039
|
-0.056
|
-0.053
|
-0.054
|
-0.071
|
|
|
(0.059)
|
(0.064)
|
(0.072)
|
(0.065)
|
(0.062)
|
(0.061)
|
(0.062)
|
(0.059)
|
(0.064)
|
(0.067)
|
(0.063)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.155***
|
0.048
|
-0.083
|
-0.013
|
0.154**
|
-0.086*
|
0.268***
|
0.225***
|
-0.053
|
0.015
|
0.012
|
|
|
(0.046)
|
(0.088)
|
(0.068)
|
(0.098)
|
(0.062)
|
(0.052)
|
(0.047)
|
(0.044)
|
(0.089)
|
(0.050)
|
(0.041)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.070
|
0.056
|
-0.013
|
-0.253*
|
|
-0.091
|
|
0.125
|
0.010
|
0.022
|
|
|
(0.000)
|
(0.116)
|
(0.131)
|
(0.147)
|
(0.131)
|
(0.000)
|
(0.066)
|
(0.000)
|
(0.119)
|
(0.065)
|
(0.059)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
456
|
456
|
456
|
455
|
435
|
434
|
434
|
429
|
432
|
2,846
|
3,947
|
|
R2
|
0.008
|
0.011
|
0.009
|
0.007
|
0.013
|
0.005
|
0.018
|
0.011
|
0.003
|
0.002
|
0.003
|
|
Adjusted R2
|
0.002
|
0.002
|
0.0002
|
-0.002
|
0.004
|
-0.004
|
0.007
|
0.002
|
-0.007
|
0.001
|
0.002
|
|
Residual Std. Error
|
0.600 (df = 452)
|
0.598 (df = 451)
|
0.599 (df = 451)
|
0.599 (df = 450)
|
0.603 (df = 430)
|
0.604 (df = 429)
|
0.600 (df = 428)
|
0.601 (df = 424)
|
0.601 (df = 427)
|
0.592 (df = 2842)
|
0.599 (df = 3943)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Curiosity 2: I explored a completely new idea or topic just for fun
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Curiosity 2: I explored a completely new idea or topic just for fun
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.069
|
-0.044
|
-0.037
|
-0.008
|
-0.037
|
0.010
|
-0.015
|
0.009
|
0.099*
|
-0.004
|
-0.023
|
|
|
(0.046)
|
(0.050)
|
(0.072)
|
(0.057)
|
(0.055)
|
(0.051)
|
(0.048)
|
(0.048)
|
(0.053)
|
(0.061)
|
(0.057)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.279***
|
0.087*
|
0.043
|
0.112***
|
0.003
|
0.011
|
0.163***
|
0.030
|
0.099
|
0.033
|
0.029
|
|
|
(0.056)
|
(0.046)
|
(0.067)
|
(0.042)
|
(0.066)
|
(0.035)
|
(0.030)
|
(0.038)
|
(0.072)
|
(0.041)
|
(0.035)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.031
|
0.014
|
-0.169
|
-0.037
|
|
-0.007
|
|
-0.246**
|
-0.021
|
-0.002
|
|
|
(0.000)
|
(0.071)
|
(0.151)
|
(0.109)
|
(0.117)
|
(0.000)
|
(0.074)
|
(0.000)
|
(0.110)
|
(0.057)
|
(0.052)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
455
|
455
|
455
|
454
|
435
|
434
|
434
|
429
|
432
|
2,846
|
3,943
|
|
R2
|
0.011
|
0.005
|
0.005
|
0.008
|
0.003
|
0.004
|
0.014
|
0.004
|
0.013
|
0.0004
|
0.001
|
|
Adjusted R2
|
0.004
|
-0.004
|
-0.004
|
-0.001
|
-0.006
|
-0.005
|
0.002
|
-0.005
|
0.004
|
-0.001
|
0.00000
|
|
Residual Std. Error
|
0.557 (df = 451)
|
0.558 (df = 450)
|
0.558 (df = 450)
|
0.557 (df = 449)
|
0.562 (df = 430)
|
0.563 (df = 429)
|
0.559 (df = 428)
|
0.563 (df = 424)
|
0.556 (df = 427)
|
0.547 (df = 2842)
|
0.558 (df = 3939)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Average curiosity
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Average curiosity
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.085*
|
-0.074
|
-0.085
|
-0.048
|
-0.015
|
-0.011
|
-0.030
|
-0.026
|
0.025
|
-0.031
|
-0.048
|
|
|
(0.045)
|
(0.046)
|
(0.064)
|
(0.053)
|
(0.052)
|
(0.046)
|
(0.044)
|
(0.045)
|
(0.054)
|
(0.056)
|
(0.052)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.223***
|
0.068
|
-0.022
|
0.051
|
0.080
|
-0.036
|
0.217***
|
0.131***
|
0.024
|
0.025
|
0.022
|
|
|
(0.045)
|
(0.060)
|
(0.058)
|
(0.060)
|
(0.052)
|
(0.036)
|
(0.030)
|
(0.036)
|
(0.079)
|
(0.041)
|
(0.033)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
-0.045
|
0.041
|
-0.089
|
-0.144
|
|
-0.045
|
|
-0.067
|
-0.004
|
0.011
|
|
|
(0.000)
|
(0.079)
|
(0.136)
|
(0.120)
|
(0.112)
|
(0.000)
|
(0.053)
|
(0.000)
|
(0.104)
|
(0.054)
|
(0.048)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
454
|
454
|
454
|
453
|
434
|
433
|
433
|
428
|
431
|
2,837
|
3,934
|
|
R2
|
0.012
|
0.009
|
0.007
|
0.008
|
0.008
|
0.006
|
0.022
|
0.009
|
0.002
|
0.001
|
0.002
|
|
Adjusted R2
|
0.006
|
0.0003
|
-0.002
|
-0.0004
|
-0.001
|
-0.004
|
0.010
|
-0.001
|
-0.007
|
0.0002
|
0.001
|
|
Residual Std. Error
|
0.495 (df = 450)
|
0.494 (df = 449)
|
0.495 (df = 449)
|
0.494 (df = 448)
|
0.502 (df = 429)
|
0.501 (df = 428)
|
0.498 (df = 427)
|
0.501 (df = 423)
|
0.499 (df = 426)
|
0.495 (df = 2833)
|
0.497 (df = 3930)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Goal setting 1: I reflect on my life goals and the type of person I want to be
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Goal setting 1: I reflect on my life goals and the type of person I want to be
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.013
|
-0.016
|
0.074*
|
0.056
|
0.027
|
0.046
|
0.048
|
0.049
|
0.069*
|
0.049
|
0.051
|
|
|
(0.033)
|
(0.041)
|
(0.042)
|
(0.043)
|
(0.043)
|
(0.036)
|
(0.034)
|
(0.038)
|
(0.041)
|
(0.047)
|
(0.044)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.201***
|
-0.006
|
0.093**
|
0.076*
|
-0.048
|
-0.029
|
0.058***
|
-0.030
|
-0.012
|
0.031
|
0.033
|
|
|
(0.037)
|
(0.031)
|
(0.040)
|
(0.045)
|
(0.055)
|
(0.023)
|
(0.018)
|
(0.034)
|
(0.066)
|
(0.041)
|
(0.032)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.090
|
-0.041
|
-0.092
|
-0.014
|
|
0.0003
|
|
-0.028
|
-0.019
|
-0.023
|
|
|
(0.000)
|
(0.062)
|
(0.098)
|
(0.059)
|
(0.066)
|
(0.000)
|
(0.093)
|
(0.000)
|
(0.084)
|
(0.053)
|
(0.047)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
455
|
455
|
455
|
454
|
435
|
434
|
434
|
429
|
432
|
2,855
|
3,943
|
|
R2
|
0.010
|
0.005
|
0.009
|
0.005
|
0.005
|
0.009
|
0.011
|
0.007
|
0.008
|
0.003
|
0.003
|
|
Adjusted R2
|
0.003
|
-0.004
|
0.0002
|
-0.003
|
-0.004
|
-0.001
|
-0.001
|
-0.003
|
-0.001
|
0.002
|
0.002
|
|
Residual Std. Error
|
0.433 (df = 451)
|
0.435 (df = 450)
|
0.434 (df = 450)
|
0.435 (df = 449)
|
0.438 (df = 430)
|
0.438 (df = 429)
|
0.438 (df = 428)
|
0.435 (df = 424)
|
0.440 (df = 427)
|
0.426 (df = 2851)
|
0.435 (df = 3939)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Goal setting 2: When planning my day, I think about how my activities are related to what I want to achieve in life
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Goal setting 2: When planning my day, I think about how my activities are related to what I want to achieve in life
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.014
|
-0.006
|
0.071
|
0.050
|
-0.009
|
0.039
|
0.051
|
0.051
|
0.089**
|
0.040
|
0.052
|
|
|
(0.041)
|
(0.043)
|
(0.056)
|
(0.047)
|
(0.047)
|
(0.039)
|
(0.042)
|
(0.043)
|
(0.039)
|
(0.048)
|
(0.045)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.118*
|
0.022
|
0.072
|
0.100***
|
-0.048
|
-0.143***
|
0.136***
|
-0.012
|
0.069
|
0.050
|
0.054*
|
|
|
(0.061)
|
(0.031)
|
(0.052)
|
(0.037)
|
(0.057)
|
(0.028)
|
(0.023)
|
(0.043)
|
(0.051)
|
(0.038)
|
(0.030)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.026
|
-0.106
|
-0.090
|
0.080
|
|
-0.095
|
|
-0.116*
|
-0.044
|
-0.054
|
|
|
(0.000)
|
(0.069)
|
(0.107)
|
(0.066)
|
(0.076)
|
(0.000)
|
(0.059)
|
(0.000)
|
(0.060)
|
(0.050)
|
(0.045)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
454
|
454
|
454
|
453
|
434
|
433
|
433
|
428
|
431
|
2,846
|
3,934
|
|
R2
|
0.008
|
0.006
|
0.009
|
0.010
|
0.006
|
0.011
|
0.013
|
0.008
|
0.016
|
0.002
|
0.003
|
|
Adjusted R2
|
0.001
|
-0.003
|
-0.0002
|
0.001
|
-0.003
|
0.002
|
0.001
|
-0.002
|
0.006
|
0.001
|
0.002
|
|
Residual Std. Error
|
0.435 (df = 450)
|
0.436 (df = 449)
|
0.435 (df = 449)
|
0.435 (df = 448)
|
0.439 (df = 429)
|
0.438 (df = 428)
|
0.438 (df = 427)
|
0.432 (df = 423)
|
0.440 (df = 426)
|
0.430 (df = 2842)
|
0.436 (df = 3930)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Average goal setting
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Average goal setting
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
0.013
|
-0.011
|
0.072
|
0.053
|
0.008
|
0.042
|
0.049
|
0.049
|
0.079**
|
0.044
|
0.051
|
|
|
(0.035)
|
(0.038)
|
(0.047)
|
(0.043)
|
(0.043)
|
(0.035)
|
(0.035)
|
(0.037)
|
(0.038)
|
(0.043)
|
(0.040)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.160***
|
0.008
|
0.082*
|
0.088**
|
-0.048
|
-0.086***
|
0.097***
|
-0.020
|
0.029
|
0.040
|
0.044
|
|
|
(0.048)
|
(0.028)
|
(0.045)
|
(0.037)
|
(0.053)
|
(0.024)
|
(0.018)
|
(0.035)
|
(0.058)
|
(0.036)
|
(0.029)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.056
|
-0.073
|
-0.094*
|
0.034
|
|
-0.046
|
|
-0.073
|
-0.032
|
-0.039
|
|
|
(0.000)
|
(0.060)
|
(0.101)
|
(0.048)
|
(0.066)
|
(0.000)
|
(0.072)
|
(0.000)
|
(0.070)
|
(0.048)
|
(0.043)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
454
|
454
|
454
|
453
|
434
|
433
|
433
|
428
|
431
|
2,846
|
3,934
|
|
R2
|
0.010
|
0.006
|
0.009
|
0.009
|
0.005
|
0.010
|
0.013
|
0.008
|
0.014
|
0.003
|
0.003
|
|
Adjusted R2
|
0.003
|
-0.003
|
0.001
|
-0.0001
|
-0.004
|
0.001
|
0.002
|
-0.001
|
0.004
|
0.002
|
0.003
|
|
Residual Std. Error
|
0.393 (df = 450)
|
0.394 (df = 449)
|
0.393 (df = 449)
|
0.394 (df = 448)
|
0.399 (df = 429)
|
0.399 (df = 428)
|
0.399 (df = 427)
|
0.393 (df = 423)
|
0.399 (df = 426)
|
0.388 (df = 2842)
|
0.395 (df = 3930)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Gratitude 1: I say “Thank you”
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Gratitude 1: I say “Thank you”
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.035
|
-0.077*
|
-0.011
|
-0.039
|
-0.067
|
0.027
|
0.039
|
0.033
|
0.080*
|
0.005
|
-0.002
|
|
|
(0.038)
|
(0.045)
|
(0.042)
|
(0.045)
|
(0.050)
|
(0.036)
|
(0.041)
|
(0.037)
|
(0.048)
|
(0.044)
|
(0.041)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.100**
|
-0.018
|
0.018
|
0.018
|
-0.056
|
-0.073**
|
0.100***
|
0.029
|
0.143***
|
0.015
|
0.022
|
|
|
(0.048)
|
(0.055)
|
(0.043)
|
(0.054)
|
(0.036)
|
(0.029)
|
(0.029)
|
(0.034)
|
(0.030)
|
(0.032)
|
(0.025)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.089
|
-0.051
|
0.045
|
0.120*
|
|
-0.065
|
|
-0.141**
|
0.005
|
0.006
|
|
|
(0.000)
|
(0.076)
|
(0.095)
|
(0.072)
|
(0.061)
|
(0.000)
|
(0.051)
|
(0.000)
|
(0.065)
|
(0.045)
|
(0.041)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
454
|
454
|
454
|
453
|
434
|
433
|
433
|
428
|
431
|
2,846
|
3,934
|
|
R2
|
0.003
|
0.006
|
0.002
|
0.004
|
0.007
|
0.010
|
0.007
|
0.008
|
0.015
|
0.0004
|
0.001
|
|
Adjusted R2
|
-0.003
|
-0.002
|
-0.007
|
-0.005
|
-0.003
|
0.0005
|
-0.004
|
-0.001
|
0.006
|
-0.001
|
-0.00003
|
|
Residual Std. Error
|
0.386 (df = 450)
|
0.386 (df = 449)
|
0.387 (df = 449)
|
0.387 (df = 448)
|
0.393 (df = 429)
|
0.392 (df = 428)
|
0.393 (df = 427)
|
0.384 (df = 423)
|
0.392 (df = 426)
|
0.385 (df = 2842)
|
0.389 (df = 3930)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Gratitude 2: I did something nice to show my appreciation
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Gratitude 2: I did something nice to show my appreciation
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.065
|
-0.083
|
-0.040
|
-0.061
|
-0.089
|
-0.006
|
-0.021
|
-0.008
|
0.052
|
-0.007
|
-0.034
|
|
|
(0.048)
|
(0.061)
|
(0.061)
|
(0.051)
|
(0.063)
|
(0.050)
|
(0.056)
|
(0.052)
|
(0.067)
|
(0.055)
|
(0.051)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.171***
|
-0.039
|
0.033
|
0.089
|
-0.082
|
-0.195***
|
0.017
|
0.138***
|
0.147**
|
-0.004
|
0.005
|
|
|
(0.046)
|
(0.048)
|
(0.063)
|
(0.071)
|
(0.070)
|
(0.043)
|
(0.044)
|
(0.039)
|
(0.068)
|
(0.039)
|
(0.031)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.102
|
0.009
|
0.051
|
0.126
|
|
0.123*
|
|
-0.109
|
0.073
|
0.071
|
|
|
(0.000)
|
(0.089)
|
(0.107)
|
(0.080)
|
(0.084)
|
(0.000)
|
(0.072)
|
(0.000)
|
(0.092)
|
(0.054)
|
(0.049)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
453
|
453
|
453
|
452
|
433
|
432
|
432
|
427
|
430
|
2,846
|
3,925
|
|
R2
|
0.007
|
0.005
|
0.004
|
0.013
|
0.007
|
0.009
|
0.009
|
0.008
|
0.009
|
0.002
|
0.002
|
|
Adjusted R2
|
0.0004
|
-0.004
|
-0.005
|
0.004
|
-0.003
|
-0.00004
|
-0.002
|
-0.001
|
-0.0002
|
0.001
|
0.001
|
|
Residual Std. Error
|
0.481 (df = 449)
|
0.482 (df = 448)
|
0.482 (df = 448)
|
0.480 (df = 447)
|
0.487 (df = 428)
|
0.486 (df = 427)
|
0.487 (df = 426)
|
0.479 (df = 422)
|
0.485 (df = 425)
|
0.479 (df = 2842)
|
0.483 (df = 3921)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|
Average gratitude
|
|
|
|
Dependent variable:
|
|
|
|
|
|
Average gratitude
|
|
|
Round 1
|
Round 2
|
Round 3
|
Round 4
|
Round 5
|
Round 6
|
Round 7
|
Round 8
|
Round 9
|
Pure Treatment
|
Perc. courses monitored
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
|
|
|
Monitoring
|
-0.050
|
-0.080
|
-0.026
|
-0.051
|
-0.078
|
0.010
|
0.009
|
0.013
|
0.065
|
-0.001
|
-0.018
|
|
|
(0.038)
|
(0.050)
|
(0.046)
|
(0.043)
|
(0.051)
|
(0.039)
|
(0.045)
|
(0.040)
|
(0.054)
|
(0.045)
|
(0.042)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Female Teacher
|
0.135***
|
-0.027
|
0.025
|
0.051
|
-0.067
|
-0.133***
|
0.059*
|
0.085***
|
0.141***
|
0.005
|
0.013
|
|
|
(0.030)
|
(0.048)
|
(0.048)
|
(0.059)
|
(0.047)
|
(0.034)
|
(0.034)
|
(0.032)
|
(0.048)
|
(0.033)
|
(0.026)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Monitoring x Female Teacher
|
|
0.095
|
-0.023
|
0.051
|
0.123*
|
|
0.028
|
|
-0.122
|
0.040
|
0.040
|
|
|
(0.000)
|
(0.072)
|
(0.098)
|
(0.065)
|
(0.064)
|
(0.000)
|
(0.055)
|
(0.000)
|
(0.075)
|
(0.046)
|
(0.041)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
452
|
452
|
452
|
451
|
432
|
431
|
431
|
426
|
429
|
2,837
|
3,916
|
|
R2
|
0.006
|
0.007
|
0.003
|
0.010
|
0.008
|
0.010
|
0.008
|
0.009
|
0.013
|
0.001
|
0.002
|
|
Adjusted R2
|
-0.0002
|
-0.002
|
-0.006
|
0.001
|
-0.002
|
0.001
|
-0.003
|
-0.0004
|
0.004
|
0.0003
|
0.001
|
|
Residual Std. Error
|
0.392 (df = 448)
|
0.392 (df = 447)
|
0.393 (df = 447)
|
0.392 (df = 446)
|
0.398 (df = 427)
|
0.398 (df = 426)
|
0.398 (df = 425)
|
0.389 (df = 421)
|
0.398 (df = 424)
|
0.391 (df = 2833)
|
0.394 (df = 3912)
|
|
|
|
Note:
|
p<0.1; p<0.05; p<0.01
|
|
|
Estimates are OLS with course fixed effects. Standard errors are clustered at the student and classs level, except for percentage monitoring and pure treatment, where they are clustered at the student level. Dependent variables are in a [-1, 1] likert scale.
|

Is the monitoring’s effect on IAT score due to selection bias?
Round 9 monitoring Heckman Models
monit_round9_models <- list()
for (heck.control in unlist(heck.controls)) {
selection_formula <- formula(paste('!skip_iat ~ monitoreo + ', heck.control))
outcome_formula <- formula(paste('iat_score ~ monitoreo + ', heck.control))
tryCatch({
heck.monit.round9.iat_score.cv <- selection(selection_formula, outcome_formula,
data = data %>% filter(round == 9))
monit_round9_models[[heck.control]] <- heck.monit.round9.iat_score.cv
# print(heck.control)
# print(summary(heck.monit.round9.iat_score.cv))
}, error=function(e){
# print(heck.control)
#print(e)
})
}
stargazer(monit_round9_models,
type = 'html',
dep.var.labels = c('IAT Score'),
covariate.labels = c('Monitoring', 'Participant - Female', 'Participant - Judge', 'Participant - Years of Tenure', 'Early Responder'),
notes=c("<span>***</span>: p<0.01; <span>**</span>: p<0.05; <span>*</span>: p<0.1",
"Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables",
"'Early responder' started the final exercises on the same weekends as the workshop."),
notes.append=F)
|
|
|
|
Dependent variable:
|
|
|
|
|
|
IAT Score
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
(12)
|
(13)
|
(14)
|
(15)
|
|
|
|
Monitoring
|
0.019
|
0.019
|
0.017
|
0.886
|
0.607
|
0.018
|
0.397
|
0.008
|
0.913
|
-0.118
|
0.017
|
0.023
|
-0.007
|
0.128
|
-0.099
|
|
|
(0.048)
|
(0.072)
|
(0.056)
|
(Inf.000)
|
(0.449)
|
(0.046)
|
(0.264)
|
(0.069)
|
(0.654)
|
(0.190)
|
(0.048)
|
(0.051)
|
(0.102)
|
(0.131)
|
(0.200)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Participant - Female
|
-0.015
|
|
|
|
1.731***
|
-0.011
|
1.257***
|
|
|
|
-0.011
|
-0.016
|
0.438***
|
|
0.449**
|
|
|
(0.039)
|
|
|
|
(0.417)
|
(0.040)
|
(0.199)
|
|
|
|
(0.041)
|
(0.039)
|
(0.095)
|
|
(0.202)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Participant - Judge
|
|
-0.0001
|
|
|
1.312***
|
|
|
0.005
|
2.092***
|
|
0.004
|
0.008
|
|
0.388***
|
0.286
|
|
|
|
(0.041)
|
|
|
(0.460)
|
|
|
(0.042)
|
(0.645)
|
|
(0.041)
|
(0.041)
|
|
(0.139)
|
(0.219)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Participant - Years of Tenure
|
|
|
-0.008
|
|
|
-0.007
|
|
-0.009
|
|
-0.011
|
-0.007
|
|
-0.124***
|
-0.050*
|
-0.033
|
|
|
|
|
(0.010)
|
|
|
(0.009)
|
|
(0.011)
|
|
(0.024)
|
(0.010)
|
|
(0.022)
|
(0.030)
|
(0.040)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Early Responder
|
|
|
|
1.945
|
|
|
1.055***
|
|
2.409***
|
0.313
|
|
-0.050
|
0.300***
|
0.442***
|
0.289
|
|
|
|
|
|
(Inf.000)
|
|
|
(0.225)
|
|
(0.615)
|
(0.192)
|
|
(0.039)
|
(0.102)
|
(0.124)
|
(0.205)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Constant
|
-0.177
|
-0.185
|
-0.146
|
-2.127
|
-2.514***
|
-0.142
|
-2.049***
|
-0.173
|
-3.358***
|
-1.116***
|
-0.144
|
-0.165
|
-0.882***
|
-0.673***
|
-1.282***
|
|
|
(0.119)
|
(0.215)
|
(0.123)
|
(Inf.000)
|
(0.294)
|
(0.095)
|
(0.126)
|
(0.161)
|
(0.412)
|
(0.095)
|
(0.106)
|
(0.133)
|
(0.136)
|
(0.183)
|
(0.235)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
370
|
370
|
370
|
370
|
370
|
370
|
370
|
370
|
370
|
370
|
370
|
370
|
370
|
370
|
370
|
|
Log Likelihood
|
-272.680
|
-272.848
|
-272.011
|
-686.564
|
-773.589
|
-271.757
|
-653.469
|
-271.939
|
-877.759
|
-555.904
|
-271.684
|
-271.784
|
-551.317
|
-445.016
|
-572.197
|
|
rho
|
0.016 (0.972)
|
0.033 (1.843)
|
0.059 (1.223)
|
1.000 (Inf.000)
|
0.990
|
0.035 (0.828)
|
1.000
|
0.334 (1.537)
|
0.990
|
1.000
|
0.054 (0.950)
|
0.030 (1.057)
|
0.996
|
1.000*** (0.001)
|
0.998
|
|
|
|
Note:
|
***: p<0.01; **: p<0.05; *: p<0.1
|
|
|
Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables
|
|
|
‘Early responder’ started the final exercises on the same weekends as the workshop.
|
Perc. courses Heckman Models
monit_ratio_models <- list(no_controls = heck.monit.ratio.iat_score)
for (heck.control in unlist(heck.controls)) {
selection_formula <- formula(paste('!skip_iat ~ monitoreo + ', heck.control))
outcome_formula <- formula(paste('iat_score ~ monitoreo + ', heck.control))
tryCatch({
heck.monit.ratio.iat_score.cv <- selection(selection_formula, outcome_formula, data = monit_perc)
monit_ratio_models[[heck.control]] <- heck.monit.ratio.iat_score.cv
# print(heck.control)
# print(summary(heck.monit.ratio.iat_score.cv))
}, error=function(e){
# print(heck.control)
# print(e)
})
}
stargazer(monit_ratio_models,
type = 'html',
dep.var.labels = c('IAT Score'),
covariate.labels = c('Monitoring', 'Participant - Female', 'Participant - Judge', 'Participant - Years of Tenure', 'Early Responder'),
notes=c("<span>***</span>: p<0.01; <span>**</span>: p<0.05; <span>*</span>: p<0.1",
"Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables",
"'Early responder' started the final exercises on the same weekends as the workshop."),
notes.append=F)
|
|
|
|
Dependent variable:
|
|
|
|
|
|
IAT Score
|
NA
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
(12)
|
(13)
|
(14)
|
(15)
|
(16)
|
|
|
|
Monitoring
|
-0.017
|
-0.014
|
0.280***
|
0.086***
|
0.232***
|
0.086***
|
0.085***
|
0.091***
|
0.084***
|
0.254
|
-0.421***
|
0.084***
|
0.091***
|
0.090***
|
-0.149***
|
0.091***
|
|
|
(0.046)
|
(Inf.000)
|
(0.015)
|
(0.021)
|
(0.057)
|
(0.020)
|
(0.018)
|
(0.019)
|
(0.027)
|
(Inf.000)
|
(0.049)
|
(0.019)
|
(0.020)
|
(0.018)
|
(0.044)
|
(0.019)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Participant - Female
|
|
0.236
|
|
|
|
-0.021
|
-0.016
|
-0.020
|
|
|
|
-0.016
|
-0.021
|
-0.016
|
|
-0.017
|
|
|
|
(Inf.000)
|
|
|
|
(0.013)
|
(0.013)
|
(0.013)
|
|
|
|
(0.014)
|
(0.013)
|
(0.013)
|
|
(0.013)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Participant - Judge
|
|
|
-0.176***
|
|
|
0.006
|
|
|
0.009
|
-0.258
|
|
0.010
|
0.016
|
|
0.219***
|
0.019
|
|
|
|
|
(0.027)
|
|
|
(0.013)
|
|
|
(0.014)
|
(Inf.000)
|
|
(0.013)
|
(0.013)
|
|
(0.043)
|
(0.013)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Participant - Years of Tenure
|
|
|
|
-0.008***
|
|
|
-0.008***
|
|
-0.009**
|
|
-0.060***
|
-0.008***
|
|
-0.008***
|
-0.019**
|
-0.008***
|
|
|
|
|
|
(0.003)
|
|
|
(0.003)
|
|
(0.003)
|
|
(0.010)
|
(0.003)
|
|
(0.003)
|
(0.008)
|
(0.003)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Early Responder
|
|
|
|
|
-0.649***
|
|
|
-0.064***
|
|
-0.311
|
0.108**
|
|
-0.066***
|
-0.063***
|
0.094**
|
-0.065***
|
|
|
|
|
|
|
(0.045)
|
|
|
(0.012)
|
|
(Inf.000)
|
(0.046)
|
|
(0.013)
|
(0.012)
|
(0.042)
|
(0.013)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Constant
|
0.111***
|
-0.599
|
0.130***
|
-0.169***
|
0.433***
|
-0.207***
|
-0.166***
|
-0.184***
|
-0.175***
|
0.330
|
-1.187***
|
-0.169***
|
-0.188***
|
-0.146***
|
-0.716***
|
-0.150***
|
|
|
(0.030)
|
(Inf.000)
|
(0.019)
|
(0.032)
|
(0.033)
|
(0.036)
|
(0.028)
|
(0.032)
|
(0.044)
|
(Inf.000)
|
(0.057)
|
(0.031)
|
(0.037)
|
(0.028)
|
(0.051)
|
(0.031)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
3,390
|
3,390
|
3,390
|
3,385
|
3,390
|
3,390
|
3,385
|
3,390
|
3,385
|
3,390
|
3,385
|
3,385
|
3,390
|
3,385
|
3,385
|
3,385
|
|
Log Likelihood
|
-3,995.410
|
-3,420.246
|
-3,368.208
|
-2,463.315
|
-4,482.514
|
-2,468.077
|
-2,458.919
|
-2,455.920
|
-2,461.761
|
-3,949.339
|
-5,704.835
|
-2,457.232
|
-2,454.309
|
-2,446.037
|
-4,023.765
|
-2,443.654
|
|
rho
|
-0.998
|
1.000 (Inf.000)
|
-1.000
|
0.026 (0.316)
|
-0.999
|
0.015 (0.300)
|
0.034 (0.238)
|
0.015 (0.267)
|
0.072 (0.460)
|
-1.000 (Inf.000)
|
1.000
|
0.045 (0.274)
|
0.016 (0.305)
|
0.034 (0.238)
|
0.998
|
0.040 (0.269)
|
|
|
|
Note:
|
***: p<0.01; **: p<0.05; *: p<0.1
|
|
|
Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables
|
|
|
‘Early responder’ started the final exercises on the same weekends as the workshop.
|
Pure control monitoring Heckman Models
monit_pure_models <- list()
for (heck.control in unlist(heck.controls)) {
selection_formula <- formula(paste('!skip_iat ~ monitoreo + ', heck.control))
outcome_formula <- formula(paste('iat_score ~ monitoreo + ', heck.control))
tryCatch({
heck.monit.pure.iat_score.cv <- selection(selection_formula, outcome_formula,
data = pure_control)
monit_pure_models[[heck.control]] <- heck.monit.pure.iat_score.cv
#print(heck.control)
#print(summary(heck.monit.pure.iat_score.cv))
}, error=function(e){
#print(heck.control)
#print(e)
})
}
stargazer(monit_pure_models,
type = 'html',
dep.var.labels = c('IAT Score'),
covariate.labels = c('Monitoring', 'Participant - Female', 'Participant - Judge', 'Participant - Years of Tenure', 'Early Responder'),
notes=c("<span>***</span>: p<0.01; <span>**</span>: p<0.05; <span>*</span>: p<0.1",
"Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables",
"'Early responder' started the final exercises on the same weekends as the workshop."),
notes.append=F)
|
|
|
|
Dependent variable:
|
|
|
|
|
|
IAT Score
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5)
|
(6)
|
(7)
|
(8)
|
(9)
|
(10)
|
(11)
|
(12)
|
(13)
|
(14)
|
(15)
|
|
|
|
Monitoring
|
-0.014
|
0.089***
|
0.086***
|
-0.357***
|
0.088***
|
0.086***
|
-0.438**
|
0.085***
|
-0.082*
|
0.093***
|
0.086***
|
-0.173**
|
0.093***
|
0.092***
|
0.092***
|
|
|
(0.109)
|
(0.027)
|
(0.020)
|
(0.067)
|
(0.024)
|
(0.019)
|
(0.189)
|
(0.022)
|
(0.047)
|
(0.020)
|
(0.020)
|
(0.067)
|
(0.019)
|
(0.021)
|
(0.019)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Participant - Female
|
1.662***
|
|
|
|
-0.024
|
-0.019
|
1.819***
|
|
|
|
-0.018
|
0.443***
|
-0.016
|
|
-0.015
|
|
|
(0.083)
|
|
|
|
(0.015)
|
(0.015)
|
(0.177)
|
|
|
|
(0.015)
|
(0.067)
|
(0.015)
|
|
(0.015)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Participant - Judge
|
|
-0.044***
|
|
|
-0.043***
|
|
|
-0.041**
|
0.289***
|
|
-0.040**
|
0.422***
|
|
-0.030*
|
-0.030*
|
|
|
|
(0.017)
|
|
|
(0.017)
|
|
|
(0.017)
|
(0.052)
|
|
(0.017)
|
(0.075)
|
|
(0.017)
|
(0.017)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Participant - Years of Tenure
|
|
|
-0.017***
|
|
|
-0.016***
|
|
-0.017***
|
|
-0.018***
|
-0.016***
|
|
-0.018***
|
-0.018***
|
-0.017***
|
|
|
|
|
(0.005)
|
|
|
(0.005)
|
|
(0.005)
|
|
(0.005)
|
(0.005)
|
|
(0.005)
|
(0.005)
|
(0.005)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Early Responder
|
|
|
|
1.126***
|
|
|
1.598***
|
|
0.209***
|
-0.075***
|
|
0.306***
|
-0.074***
|
-0.071***
|
-0.071***
|
|
|
|
|
|
(0.053)
|
|
|
(0.184)
|
|
(0.049)
|
(0.015)
|
|
(0.070)
|
(0.015)
|
(0.015)
|
(0.015)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Constant
|
-1.450***
|
-0.214***
|
-0.141***
|
-1.271***
|
-0.205***
|
-0.136***
|
-2.319***
|
-0.134***
|
-0.706***
|
-0.109***
|
-0.129***
|
-1.053***
|
-0.105***
|
-0.105***
|
-0.101***
|
|
|
(0.068)
|
(0.046)
|
(0.029)
|
(0.030)
|
(0.042)
|
(0.029)
|
(0.134)
|
(0.031)
|
(0.030)
|
(0.030)
|
(0.030)
|
(0.047)
|
(0.029)
|
(0.031)
|
(0.030)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Observations
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
2,432
|
|
Log Likelihood
|
-4,097.387
|
-1,792.339
|
-1,779.133
|
-3,515.098
|
-1,790.687
|
-1,777.146
|
-5,010.240
|
-1,774.010
|
-2,729.838
|
-1,766.392
|
-1,771.923
|
-3,322.238
|
-1,764.728
|
-1,762.832
|
-1,760.989
|
|
rho
|
1.000*** (0.00004)
|
0.024 (0.393)
|
0.053 (0.249)
|
1.000* (0.606)
|
0.032 (0.345)
|
0.044 (0.219)
|
1.000 (1.823)
|
0.074 (0.279)
|
1.000*** (0.153)
|
0.052 (0.245)
|
0.054 (0.234)
|
1.000*** (0.0001)
|
0.041 (0.217)
|
0.065 (0.263)
|
0.048 (0.226)
|
|
|
|
Note:
|
***: p<0.01; **: p<0.05; *: p<0.1
|
|
|
Estimates are Heckman-style selection models. Every column shows coefficients for different combinations of control variables
|
|
|
‘Early responder’ started the final exercises on the same weekends as the workshop.
|
After exploring all possible combinations of control variables: "female", "level", "judge", "years_tenure", "early_participant", on the 9th round of monitoring, the percentage f monitored courses and the pure treatment especifications we found three models with significant rhos (none of the models for the percentage of courses monitored had a significant rho):
#formula <- as.formula(paste0(outcome, ' ~ monitoreo | idcurso | 0 | aula'))
out_list <- list()
for (r in 1:9) {
tryCatch({
model <- selection(!skip_iat ~ monitoreo + female + level_2 + judge + years_tenure + early_participant, iat_score ~ monitoreo + female + level_2 + level_3 + judge + years_tenure + early_participant, data = data %>% filter(round == r))
out_list[[sprintf('Round %i', r)]] <- model
}, error=function(e){
})
}
tryCatch({
out_list[['Pure Treatment']] <- selection(!skip_iat ~ monitoreo + female + level_2 + judge + years_tenure + early_participant,
iat_score ~ monitoreo + female + level_2 + judge + years_tenure + early_participant, data = pure_control)
}, error=function(e){
})
tryCatch({
out_list[['Perc. courses monitored']] <- selection(!skip_iat ~ monitoreo + female + level_2 + level_3 + judge + years_tenure + early_participant,
iat_score ~ monitoreo + female + level_2 + level_3 + judge + years_tenure + early_participant, data = monit_perc)
}, error=function(e){
})
NULL
# stargazer(out_list,
# type = 'html',
# dep.var.labels = c('IAT Score'),
# notes=c("<span>***</span>: p<0.01; <span>**</span>: p<0.05; <span>*</span>: p<0.1",
# "Estimates are Heckman-style selection models.",
# "'Early responder' started the final exercises on the same weekends as the workshop."),
# notes.append=F)
Lee (2009) approach
Lee (2009) proposes an estimator for treatment effect bounds under the assumptions of random assignment of treatment and monotonicity. We use this approach since the monitoring intervention increased the likelihood of the students skipping the IAT score sections, potentially creating bias in the IAT score estimates. We find lower bounds different from zero from round 6 onwards, suggesting an incremental effect of the monitoring intervention.
## Estimation of Lee Bounds for IAT scores
# Load Utils (leebounds functions)
source(paste0(github,"utils/utils_leebounds.R"))
# Find mean of IAT scores for those in the treatment group (monitoring)
mean_treatment_group <- data %>%
filter(monitoreo == 1) %>%
group_by(monitoreo) %>%
summarize(mean_treatment = mean(iat_score, na.rm = TRUE))
mean_treatment_group <- mean_treatment_group[[2]]
# Preprocess dataset for rounds
leebound_rounds.list <- data %>%
select("dni", "round", "monitoreo", "skip_iat", "iat_score", "female", "level_2", "level_3", "judge", "years_tenure", "early_participant") %>%
filter(!is.na(skip_iat)) %>%
mutate(
selection = !(skip_iat),
#iat_score = case_when(
# iat_score == iat_score ~ iat_score,
# selection == TRUE ~ mean_treatment_group),
years_tenure = replace(years_tenure, is.na(years_tenure), 0),
vars_deleted = is.na(iat_score) & skip_iat == 0) %>%
rename(treat = monitoreo, outcome = iat_score) %>%
filter(vars_deleted == FALSE) %>%
group_split(round) %>%
as.list() %>%
map(pivot_wider)
# Define Control vars
control_vars <- c("female", "judge")
#control_vars <- c("female", "level_2")
# Linear model fit
leebound_rounds.list <- lapply(leebound_rounds.list, function(x){
lm.fit_x<-lm(form=as.formula(paste0("outcome~",paste0(control_vars,collapse="+")) ),
data=x[x$selection==1,])
x$predicted_iat_score<-predict(lm.fit_x,x)
x$predicted_iat_group<-sapply(x$predicted_iat_score,Lee_grouping)
return(x)
})
# Lee bounds -------------------------------------------------------------------
# Lee bounds without controls
estimate <- numeric()
for(i in 1:9) {
out_est <- GetBounds(leebounds(leebound_rounds.list[[i]]))
estimate <- c(estimate, out_est)
}
# Lee bounds without controls
estimate_r1 <-GetBounds(leebounds(leebound_rounds.list[[1]]))
estimate_r2 <-GetBounds(leebounds(leebound_rounds.list[[2]]))
estimate_r3 <-GetBounds(leebounds(leebound_rounds.list[[3]]))
estimate_r4 <-GetBounds(leebounds(leebound_rounds.list[[4]]))
estimate_r5 <-GetBounds(leebounds(leebound_rounds.list[[5]]))
estimate_r6 <-GetBounds(leebounds(leebound_rounds.list[[6]]))
estimate_r7 <-GetBounds(leebounds(leebound_rounds.list[[7]]))
estimate_r8 <-GetBounds(leebounds(leebound_rounds.list[[8]]))
estimate_r9 <-GetBounds(leebounds(leebound_rounds.list[[9]]))
# Lee bounds with controls
estimates_sharp_r1 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[1]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r2 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[2]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r3 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[3]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r4 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[4]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r5 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[5]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r6 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[6]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r7 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[7]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r8 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[8]],group_name="predicted_iat_group",treat_helps = FALSE))
estimates_sharp_r9 <-GetBounds(Lee_sharp_bounds(leebound_rounds.list[[9]],group_name="predicted_iat_group",treat_helps = FALSE))
# Lee Bounds for Pure control and Monit Percentage -----------------------------
## Pure control
# Preprocess dataset
leebound_pure_control <- pure_control %>%
select("dni", "monitoreo", "skip_iat", "iat_score", "female", "level_2", "level_3", "judge", "years_tenure", "early_participant") %>%
filter(!is.na(skip_iat)) %>%
mutate(
selection = !(skip_iat),
#iat_score = case_when(
#iat_score == iat_score ~ iat_score,
#selection == TRUE ~ mean_treatment_group),
years_tenure = replace(years_tenure, is.na(years_tenure), 0),
vars_deleted = is.na(iat_score) & skip_iat == 0,
weights = 1) %>%
filter(vars_deleted == FALSE) %>%
rename(treat = monitoreo, outcome = iat_score)
# Linear model fit
lm.fit_pure_control <-lm(form=as.formula(paste0("outcome~",paste0(control_vars,collapse="+")) ),
data=leebound_pure_control[leebound_pure_control$selection==1,])
leebound_pure_control$predicted_iat_score<-predict(lm.fit_pure_control,leebound_pure_control)
leebound_pure_control$predicted_iat_group<-sapply(leebound_pure_control$predicted_iat_score,Lee_grouping)
# Lee bounds
estimate_pure_control <-GetBounds(leebounds(leebound_pure_control))
# Lee bounds with controls
estimate_sharp_pure_control <-GetBounds(Lee_sharp_bounds(leebound_pure_control,group_name="predicted_iat_group",treat_helps = FALSE))
## Monit_percentage
# Preprocess dataset
leebound_monit_perc <- monit_perc %>%
select("dni", "monitoreo", "skip_iat", "iat_score", "female", "level_2", "level_3", "judge", "years_tenure", "early_participant") %>%
filter(!is.na(skip_iat)) %>%
mutate(
selection = !(skip_iat),
#iat_score = case_when(
# iat_score == iat_score ~ iat_score,
# selection == TRUE ~ mean_treatment_group),
years_tenure = replace(years_tenure, is.na(years_tenure), 0),
vars_deleted = is.na(iat_score) & skip_iat == 0,
weights = 1) %>%
filter(vars_deleted == FALSE) %>%
rename(treat = monitoreo, outcome = iat_score)
# Linear model fit
lm.fit_monit_perc <-lm(form=as.formula(paste0("outcome~",paste0(control_vars,collapse="+")) ),
data=leebound_monit_perc[leebound_monit_perc$selection==1,])
leebound_monit_perc$predicted_iat_score<-predict(lm.fit_monit_perc,leebound_monit_perc)
leebound_monit_perc$predicted_iat_group<-sapply(leebound_monit_perc$predicted_iat_score,Lee_grouping)
# Lee bounds
estimate_monit_perc <-GetBounds(leebounds(leebound_monit_perc))
# Lee bounds with controls
estimate_sharp_monit_perc <-GetBounds(Lee_sharp_bounds(leebound_monit_perc,group_name="predicted_iat_group",treat_helps = FALSE))
# Generalized lee bounds -------------------------------------------------------
N_rep=500
ci_alpha=0.05
quantile_grid_size=0.01
# Pure control ----
form_nonmonotone<-as.formula(paste0("selection~(treat)*(", paste0(control_vars,collapse="+"),")"))
min_iat=min(leebound_pure_control$outcome[leebound_pure_control$selection==1])
max_iat=max(leebound_pure_control$outcome[leebound_pure_control$selection==1])
glm.fit<-glm(form=form_nonmonotone,data= leebound_pure_control,family="binomial")
res<-predict_selection(glm.fit, leebound_pure_control)
s.hat=data.frame(s.0.hat=res$s.0.hat, s.1.hat=res$s.1.hat)
p.0.star<-s.hat$s.0.hat/s.hat$s.1.hat
inds_helps<-(p.0.star<=1)
inds_hurts<-(!inds_helps)
# Estimates
estimates_g_pure_control<-GetBounds(leebounds_wout_monotonicity(leebound_pure_control,p.0.star))
bounds_g_pure_control<-main_bb(function_name=leebounds_wout_monotonicity,p.0.star=p.0.star,mydata=leebound_pure_control[,c("treat","selection","outcome","weights")],N_rep=N_rep)
## confidence region for identified set
CR_g_pure_control<-compute_confidence_region(bounds_g_pure_control,estimates_g_pure_control, ci_alpha=ci_alpha )
# Monit_perc ----
form_nonmonotone<-as.formula(paste0("selection~(treat)*(", paste0(control_vars,collapse="+"),")"))
min_iat=min(leebound_monit_perc$outcome[leebound_monit_perc$selection==1])
max_iat=max(leebound_monit_perc$outcome[leebound_monit_perc$selection==1])
glm.fit<-glm(form=form_nonmonotone,data= leebound_monit_perc,family="binomial")
res<-predict_selection(glm.fit, leebound_monit_perc)
s.hat=data.frame(s.0.hat=res$s.0.hat, s.1.hat=res$s.1.hat)
p.0.star<-s.hat$s.0.hat/s.hat$s.1.hat
inds_helps<-(p.0.star<=1)
inds_hurts<-(!inds_helps)
# Estimates
estimates_g_monit_perc<-GetBounds(leebounds_wout_monotonicity(leebound_monit_perc,p.0.star))
bounds_g_monit_perc<-main_bb(function_name=leebounds_wout_monotonicity,p.0.star=p.0.star,mydata=leebound_monit_perc[,c("treat","selection","outcome","weights")],N_rep=N_rep)
## confidence region for identified set
CR_g_monit_perc<-compute_confidence_region(bounds_g_monit_perc,estimates_g_pure_control, ci_alpha=ci_alpha )
# Generalized estimates
source(file.path("/Users/bran/Dropbox/WB/AMAG1/leebounds/generalized_lee.R"))
### Table with results ---------------------------------------------------------
# Table
ht <- hux(
Round = c('Round 1', 'Round 2', 'Round 3', 'Round 4', 'Round 5', 'Round 6', 'Round 7', 'Round 8', 'Round 9', 'Pure Treatment', 'Perc. courses monitored'),
Basic = c(estimate_r1, estimate_r2, estimate_r3, estimate_r4, estimate_r5, estimate_r6, estimate_r7, estimate_r8, estimate_r9, estimate_pure_control, estimate_monit_perc),
Covariate_Based = c(estimates_sharp_r1, estimates_sharp_r2, estimates_sharp_r3, estimates_sharp_r4, estimates_sharp_r5, estimates_sharp_r6, estimates_sharp_r7, estimates_sharp_r8, estimates_sharp_r9, estimate_sharp_pure_control, estimate_sharp_monit_perc),
Generalized = c(estimates_g_r1, estimates_g_r2, estimates_g_r3, estimates_g_r4, estimates_g_r5, estimates_g_r6, estimates_g_r7, estimates_g_r8, estimates_g_r9, estimates_g_pure_control, estimates_g_monit_perc),
add_colnames = TRUE)
ht <- add_footnote(ht, "N between 434 (round 1) and 560 (pure control) observations. Each panel reports estimated lower and upper bounds. Column (1) reports basic Lee Bounds. Column (2) reports covariate-based Lee bounds. Covariates include: female, level, judge. Column (3) reports Generalized Lee Bounds from Semenova (2021)")
ht %>%
set_bold(1, everywhere) %>%
set_bottom_border(1, everywhere) %>%
set_align(everywhere, 2, 'right') %>%
set_lr_padding(10) %>%
set_width(0.7) %>%
set_number_format(col= "Round", value = 0)
| Round | Basic | Covariate_Based | Generalized |
| Round 1 | [-0.002, 0.138] | [-0.003, 0.139] | [-0.012, 0.146] |
| Round 2 | [0.001, 0.141] | [-0.002, 0.143] | [-0.007, 0.147] |
| Round 3 | [-0.003, 0.143] | [-0.004, 0.144] | [-0.009, 0.148] |
| Round 4 | [-0.002, 0.144] | [-0.003, 0.145] | [-0.008, 0.149] |
| Round 5 | [0.003, 0.145] | [0.001, 0.15] | [0.001, 0.147] |
| Round 6 | [0.019, 0.138] | [0.023, 0.151] | [0.019, 0.138] |
| Round 7 | [0.022, 0.155] | [0.026, 0.168] | [0.022, 0.155] |
| Round 8 | [0.03, 0.158] | [0.033, 0.168] | [0.03, 0.158] |
| Round 9 | [-0.039, 0.077] | [-0.03, 0.088] | [-0.042, 0.079] |
| Pure Treatment | [0.001, 0.171] | [-0.004, 0.18] | [0.001, 0.171] |
| Perc. courses monitored | [-0.006, 0.164] | [-0.003, 0.161] | [-0.006, 0.164] |
| N between 434 (round 1) and 560 (pure control) observations. Each panel reports estimated lower and upper bounds. Column (1) reports basic Lee Bounds. Column (2) reports covariate-based Lee bounds. Covariates include: female, level, judge. Column (3) reports Generalized Lee Bounds from Semenova (2021) |